Vestergaard Frandsen A/S (now Called MVF 3Aps) v Bestnet Europe Ltd & Ors

In April 2009 Arnold J (‘the Judge’) handed down judgment in which he held that the Defendants had misused the Claimants’ confidential information: [2009] EWHC 657 (Ch) (‘the Liability Judgment’). That followed a 16 day trial earlier in 2009 (‘the Liability Trial’). He subsequently ordered that there be an inquiry as to the compensation to which the Claimants are entitled as a result of that misuse. This judgment follows the hearing of that inquiry as to damages (‘the Inquiry Hearing’). The parties are very far apart on the quantum of the damages to which Claimants are entitled. The Claimants say their claim is worth many millions of pounds and the Defendants say it is worth only a few thousand pounds. There have been a number of judgments handed down in these proceedings both by Arnold J and by the Court of Appeal. The full background to the case is set out in those judgments and I set out here only those facts which are necessary to understand the issues that arose at the Inquiry Hearing.

The mosquito nets

The action concerns secret information used in the manufacture of long lasting insecticidal mosquito bed nets mainly sold in areas of the world where malaria is endemic. Bed nets, as the name suggests, are mosquito nets that are erected over a bed to protect people from mosquitoes whilst they sleep. The nets made by the Defendants are formed from knitted polyethylene yarn. The yarn contains an insecticide called deltamethrin. The insecticide migrates to the surface of the yarn and is picked up through the feet of the insect which lands on the net and the insect dies soon after. When the nets are washed, the surface insecticide is removed. However, because the migration of insecticide from within the yarn is continuous, more insecticide comes to the surface, regenerating the insecticidal properties of the net. The control of the rate of migration of the insecticide is important. If the insecticide migrates too quickly to the surface of the net after washing, the reservoir of insecticide within the yarn will deplete too rapidly, reducing the useful life of the net. If the insecticide migrates too slowly, the regeneration time will be too long and the net will not be effective again soon after washing. Various additives which are also incorporated in the yarn affect the rate of migration of insecticide to the surface. The confidential information in issue in the action concerned the choice of additives and in particular their relative proportions, referred to in the evidence and judgments as “recipes” or “formulae”.

Nets can be made either of polyethylene (PE) yarn or polyester (PES). The manufacture process is different for the different materials. In this case we are mainly concerned with the Defendants’ bed net product called Netprotect. That is made with PE yarn. The yarn is made by putting pellets of polyethylene into an extruder which melts the pellets and pushes out the thin yarn which is then knitted up into netting and sewn into the appropriate shape to suspend over a bed. The insecticide and any additives are incorporated into the yarn by adding to the pure polyethylene pellets a ‘masterbatch’ of PE pellets containing a certain percentage of insecticide or additive. In the extruder, the ordinary pellets and the different masterbatch pellets are melted and combined so that the final yarn has an even distribution of insecticide and additive.

The additives with which we are most concerned in this case are:

Since the recipes used are still secret, much of the Inquiry Hearing was held in private. The Liability Judgment redacted the names of the additives and the recipes discussed. I shall use the same substitute lettering as was used in the Liability Judgment, namely:

The bioassaying and testing of nets

When a recipe for a long lasting insecticidal net (‘LLIN’) is being evaluated, the samples of net made up using that recipe must undergo a number of tests. The tests are primarily bioassay tests in which mosquitoes are exposed to the sample of the net for a set time to see how many of them die or are disabled as a result. The first tests are aimed at working out the regeneration time of the nets, that is how long it takes the net, after washing, to regenerate its insecticidal properties so as to kill the required number of mosquitoes within a given time. Regeneration time for LLINs is typically a day or more. Once the regeneration time has been established, promising samples undergo wash tests. The nets are washed multiple times, allowing the net to regenerate between each wash and then they are bioassayed either after each wash/regeneration or after a number of washes, again to see whether they still kill enough mosquitoes. This determines how many washes the net can undergo before it effectively runs out of insecticide.

In addition, a sample net will undergo chemical composition tests, for example to work out how much of insecticide or additives have been lost in the extrusion process at the factory before any wash testing or how much insecticide is left in the net after the final wash testing has been completed. A sample will also be tested to establish fabric integrity and toxicology tests will be carried out to see what effect the net might have on the people using it.

Some of these tests can be conducted in a laboratory. The laboratory that has featured in these proceedings is that operated by Laboratoire de Lutte contre les Insectes nuisibles in Montpellier in France (‘the Montpellier Lab’). The wash tests are usually conducted as field trials using experimental huts in places where malaria is endemic. The test site that is most relevant to this case is the Centre Muraz in Burkina Faso.

Approval of LLINs by WHO

The World Health Organisation operates a system for granting approval to LLINs after subjecting them to its own testing regime. The testing and grant of approval is carried out under the WHO Pesticide Evaluation Scheme (‘WHOPES’). WHO has a specification for LLINs which it approves; a net must achieve a regeneration time of not more than 3 days and it must have a minimum insecticidal effect after 20 washes and 3 years of recommended use under field conditions. There are three phases to WHOPES testing:

Following the successful completion of WHOPES I and WHOPES II testing, WHOPES will issue an interim approval for the net. Many customers will only buy nets that have this interim WHOPES approval.

Before WHO accept an application for testing, the applicant must provide a package of information including a draft technical specification, a risk assessment and related data. WHO does not have the resources to evaluate every product proposed by a manufacturer, even though the manufacturer has to pay a fee for the testing. So it will consider the material provided before deciding whether to go ahead with the WHOPES evaluation process. Products submitted for WHOPES evaluation should have reached a stage where they are capable of manufacture and are suitable for supply to end users. A full report of any laboratory or field tests on the proposed product should be included with the package of information supplied. In general WHOPES has one meeting per year, usually in December, at which it considers whether to grant interim approval to a product that has undergone evaluation. The timing of the meeting is in part at least dictated by when the data for all products being tested in the pipeline is ready.

The market for sales of LLINs

It is estimated that each year around 1 million children under the age of 5 die from malaria-related diseases, mainly in Africa, Latin and South America and South-East Asia. The main market for bed nets, particularly during the years when the events of this case took place, is sub-Saharan Africa. In 1998, in order to provide a coordinated global approach to fighting malaria, a project called Roll Back Malaria was launched by WHO, the United Nations Children’s Fund (UNICEF), the United National Development Programme and the World Bank. One of the key methods that has been employed to combat malaria is the provision of insecticidal mosquito nets. The customers who buy the nets are not, therefore, the ultimate users of the nets but rather these charities or governmental organisations concerned with preventing malaria. They buy many hundreds of thousands of nets a year, usually inviting manufacturers to tender for the supply of hundreds of thousands of nets.

The parties

The Claimants, to whom I shall refer collectively as “VF”, are part of a group of companies which was founded in Denmark in 1957. The history of the group is set out in paragraphs 8 onwards of the Liability Judgment. In about 1986, VF began to supply emergency relief products and in about 1993-94 they began to manufacture and sell PE tarpaulins for use in refugee camps. At about the same time, VF started supplying disease control products, beginning with a tse-tse fly trap which was made from a woven PES fabric impregnated with deltamethrin insecticide.

The Fifth Defendant, Torben Holm Larsen, was employed as VF’s Head of Production from 1 November 2000. He resigned from that post on 18 May 2004 and worked out half his notice period, leaving the company finally on 31 August 2004. Also engaged by VF on 1 November 2000 was Trine Sig who was the Sixth Defendant at the time of the Liability Judgment but who was ultimately held not to be liable to VF for breach of confidence. Mrs Sig was employed by VF initially as a sales and marketing assistant and later as Regional Sales Manager for Europe and Latin America. She handed in her notice on 17 June 2004, and it was agreed that her employment should terminate on the same day. On 3 August 2004 Mr Larsen and Mrs Sig incorporated a Danish company called Intection A/S, later Intection ApS (“Danish Intection”), to compete with VF in the field of LLINs. Danish Intection ceased to trade in about October 2005 after proceedings had been brought against it by VF in Denmark.

The corporate Defendants are all English companies which were set up at about the time that Danish Intection ceased trading. The First Defendant, Bestnet Europe Ltd (“Bestnet”) is the principal corporate defendant. Its shareholders include a company called Intelligent Insect Control SARL (“IIC”). Mr Larsen and Mrs Sig were appointed as directors of Bestnet in 2008. The Second Defendant, Intection Ltd, was originally set up to perform the role now performed by Bestnet. Its activities, if any, were very short-lived and it is now in liquidation. The Third Defendant, 3T Europe Ltd is the corporate vehicle through which Mr Larsen and Mrs Sig provide their services to Bestnet. The Fourth Defendant, Intelligent Insect Control Ltd, is a company owned by Dr Ole Skovmand, his wife and two relatives of theirs. It does not appear to have traded.

It was not suggested during the Inquiry Hearing that I need to distinguish between the Defendants in terms of their potential liability to pay to VF such compensation as I determine should be paid. In this judgment therefore I will do as Arnold J did in the Liability Judgment and refer generally to “the Defendants” without distinguishing between them unless it is necessary to do so.

Dr Ole Skovmand is central to the issues raised in this inquiry as he was central to the issues of liability. Dr Skovmand’s qualifications and background are described in paragraphs 20 onwards of the Liability Judgment. He has for many years worked on evaluating and testing insecticides on flies and mosquitoes. His work included testing insecticides incorporated in polymers, running bioassays and testing polymers. It is common ground that from April 1999 until he resigned on 7 December 2004 Dr Skovmand spent a significant proportion of his time working as a consultant for VF, initially under an informal oral arrangement and subsequently pursuant to a written agreement between VF and IIC. I also include Dr Skovmand and IIC in references to “the Defendants” unless the context indicates to the contrary.

The development of Netprotect

The history of Dr Skovmand’s work for VF on insecticidal materials and the start of the rival business set up by him together with Mrs Sig and Mr Larsen is described in detail in the Liability Judgment. Dr Skovmand was engaged by VF in late 1998 to work on an insecticidal mosquito bed net made of polyester rather than polyethylene. VF’s polyester LLIN is called PermaNet.

In late 2000 Dr Skovmand also started working for VF on developing an insecticidal polyethylene tarpaulin comprising different layers of polymer also incorporating insecticide. At about the same time, VF decided to develop a PE net incorporating insecticide. The PE net project was aimed at producing both a net for use as an agricultural fence to protect animals and crops against insects and at producing a mosquito bed net.

In the course of his work, Dr Skovmand established a database for VF referred to as the Fence database. This set out in a number of columns all the important details of the samples of nets made to various different recipes including the results of any tests carried out on those samples. On 2 and 3 April 2002 VF’s manufacturer of the PE nets in India produced samples of PE nets containing various concentrations of deltamethrin, [Additives A, B and C]. These became samples numbered 7 – 20 in the Fence database. In early July 2002, samples 7 – 20 were sent for wash resistance bioassay tests and the results recorded in the Fence database. The results of the bioassays on the nets appeared very encouraging. Although Arnold J found that in fact the results were over-optimistic, they did indicate whether any and if so which of the samples were promising for use as a bed net. Further work was done during late 2002 and spring 2003 on making and testing different samples for PE nets. The Judge found that in April 2003 VF’s development of PE bed nets and Fence were still proceeding in parallel.

In January 2004 PermaNet (that is VF’s polyester bed net) passed WHOPES II testing and thereby achieved an interim recommendation from WHOPES.

Arnold J held that although VF did not abandon the idea of producing a PE bed net, it was not a priority as from 2004. Thereafter production and testing of PE yarn by VF was focussed on producing a fence rather than on bed nets. Further work was done by Dr Skovmand on revising recipes and experimenting with different extrusion temperatures. In March 2004 a further recipe was devised and used to make samples in April, May and June 2004. The results of the tests on these samples indicated the importance of [Additive A] in [REDACTED]. Dr Skovmand’s work on PE nets for VF continued during 2004 and in December 2004, Dr Skovmand resigned his consultancy with VF.

Dr Skovmand then started working with Mr Larsen and Mrs Sig for Danish Intection and latterly Bestnet. In that work Dr Skovmand focused on producing a polyethylene bed net, the product which became Netprotect. In the course of this work, he set up a new database in broadly the same format as the Fence database. In the Netprotect database Dr Skovmand recorded all the different samples that were made and the test results for them.

The Netprotect bed net was launched at a Netmark conference in Kenya in October 2005 and at the MIM conference in Cameroon in November 2005. There were some sales in the months after launch before the nets obtained interim WHOPES approval. Precisely how many such sales there were was one of the points in issue at the Inquiry Hearing.

The Netprotect nets were manufactured for Bestnet by a factory in India owned by Mr M. Sivasamy trading as Siva Enterprises (‘Siva’). The factory is in Karur which is in the South Indian state of Tamil Nadu. The masterbatches containing the insecticide and the different additives were made for Bestnet by Alok Masterbatches Ltd (‘Alok’). Masterbatch was ordered by Siva and supplied by Alok directly to Siva’s factory.

The Defendants submitted an application for WHOPES approval for Netprotect in March 2006. The product passed WHOPES I and WHOPES II tests and was granted WHOPES interim approval in March 2008, following a meeting of the relevant WHOPES Committee held in December 2007.

The Netprotect product proved very successful and since it was launched about 57 million Netprotect nets have been sold. Most of the nets supplied by Bestnet were sold on a retail basis directly to the charity or other aid organisation needing them. However, there were some wholesale supplies of Netprotect, in particular through a wholesaling company called Syngenta which marketed the net under the brand name Icon Life.

In September 2013, WHOPES approval was withdrawn from Netprotect because it was shown that it was not sufficiently effective over a prolonged period of use in the field. Since then, Bestnet has sold a different bed net product which it is licensed to sell by a company called Shobika Industries (‘Shobika’). September 2013 therefore marks the end point of VF’s claim for damages from sales of nets by Bestnet since VF makes no claim in respect of sales by Bestnet of Shobika licensed products.

The findings of misuse of confidential information in the Liability Judgment

The main finding of Arnold J in the Liability Judgment was that Dr Skovmand had misused the confidential information in VF’s Fence database as the starting point for his work on Netprotect for the Defendants. The Judge set out in detail his findings as to how the Netprotect net had been developed after Dr Skovmand left VF. He considered the various ways in which it was suggested that Dr Skovmand had arrived at the first recipes in the Netprotect database without using the Fence database and rejected them.

His key finding was that the recipe used to make the first nets was based on the formula in the Fence database. This recipe, which I shall refer to as the First Formula, was the [REDACTED] recipe, that is to say, it used [Figure 1] parts [Additive A], [Figure 2] parts [Additive B] and [Figure 1] parts [Additive C]. Although the [First Formula] recipe did not appear in the Fence database, there were similar recipes in the Fence database and the results of tests of net samples made to those recipes had, the Judge found, provided a good starting point for Dr Skovmand’s work on the competing net for Bestnet.

Another key finding of the Liability Judgment, however, was that the recipe used for the nets which were sent to WHOPES in March 2006 and on which the WHOPES evaluation tests were carried out were not made using the First Formula. The samples which were ultimately given WHOPES interim approval were made using a different recipe or formula known as Recipe [REDACTED], that is using [Figure 3] parts [Additive A], [Figure 2] parts [Additive B] and [Figure 4] parts [Additive C]. I refer to this recipe as the Later Formula. Arnold J said at paragraph 672:

“the misuse of VF’s trade secrets I have found was merely the starting point for a substantial program of further development which resulted in a formulation which is different from any of VF’s recipes in a number of respects, and in particular (i) the polymer composition (at least in the case of the sample submitted for WHOPES II evaluation), (ii) the inclusion of [Additive L] and (iii) the inclusion of [Additive M]. In addition, a substantial period of time has elapsed since then.”

There were subsequent variations in the recipes used to make nets between the time that application for WHOPES approval was submitted in March 2006 and September 2013 when Bestnet stopped selling Netprotect. These changes were to other aspects of the recipe, such as the polymer mix and other additives. In particular, the samples submitted for Phase II of WHOPES testing were a different version of the Later Formula recipe from the samples submitted for Phase I of WHOPES testing. However, all the nets made after the submission of the samples for Phase I WHOPES testing used deltamethrin as the insecticide and the three main additives in the proportions used in the Later Formula. Neither side at the Liability Trial placed particular reliance on differences between the recipes used after WHOPES interim approval. VF argued that all the nets sold before September 2013 regardless of recipe should be treated in the same way and compensation paid in relation to them. The Defendants drew the line between nets made to the First Formula on the one hand and all later nets on the other hand.

Proceedings between the parties

There have been various sets of proceedings between these parties in different jurisdictions. Only some of them are relevant to this Inquiry.

Following the Liability Judgment, Arnold J handed down a further judgment on 26 June 2009 after hearing submissions on various consequential matters (‘the Remedies Judgment’). The most relevant part of the Remedies Judgment for the purposes of the Inquiry Hearing was his decision about the scope of the injunctive relief that should be granted. The main dispute was whether the injunction granted should prevent the Defendants selling nets made to the Later Formula. He decided that it should not because he had found that the Later Formula did not amount to a misuse of VF’s confidential information but was a formula derived from that misuse. It would therefore be disproportionate to prohibit the sales of the then current Netprotect product.

On 2 July 2009 Arnold J ordered, amongst other things, that there be an inquiry as to damages or equitable compensation for breach of confidence or, at VF's option, an account of Bestnet's profits. On 8 December 2009 VF elected for an inquiry as to damages or equitable compensation rather than an account of profits. Pleadings in the enquiry were subsequently exchanged.

Both sides challenged various findings of the Liability and Remedies Judgments before the Court of Appeal. Before the hearing of the main appeal, the Court of Appeal remitted certain factual questions to Arnold J pursuant to CPR 52.10(2)(b). This resulted in a further judgment on 7 March 2011 in which Arnold J answered those questions (‘The Remitted Questions Judgment’).

The Court of Appeal gave judgment dealing with an appeal, a cross-appeal and a request for permission to appeal from the four judgments that had, at that stage been handed down by Arnold J in these proceedings. The judgment of the Court of Appeal was handed down on 20 April 2011. Jacob LJ (with whom Jackson LJ and Sir John Chadwick agreed) held that:-

VF appealed to the Supreme Court against the Court of Appeal’s decision to overturn the judgment against Mrs Sig. That appeal was unsuccessful: see [2013] UKSC 31.

Returning to the progress of the inquiry as to damages, VF’s Amended Particulars of Claim sought damages in respect of (i) sales of products made to the First Formula when the product was launched in October 2005; (ii) products made to the Later Formula in respect of which the Defendants obtained WHOPES I and II approval in March 2008; (iii) products sold “… under or by reference to or in reliance upon the WHOPES I and WHOPES II evaluations obtained by the Defendants” in respect of those earlier Netprotect products; and (iv) products manufactured in accordance with any formulation derived in whole or in part from the misuse by the Defendants of the Claimants' trade secrets as held at trial. VF's primary case was that it was entitled to lost profits on sales that VF would have made in the absence of Bestnet on the market, or alternatively a royalty on all other sales of products, whether those products were made using the confidential information or were merely "derived". An alternative case was also pleaded based on a royalty on all Bestnet's sales of unlawful products.

The Defendants brought an application to strike out much of what was claimed in the Amended Particulars of Claim. In effect they sought to eliminate altogether from the scope of the Inquiry Hearing anything other than sales of the Netprotect product made to the First Formula. They argued that since Arnold J had held that the manufacture and sale of the mosquito nets made in accordance with the formulation submitted for WHOPES evaluation did not amount to misuse of confidential information, VF could not claim damages or equitable compensation in respect of the impact on VF of sales of those products.

The Defendants’ strike out application was dismissed by Arnold J in a judgment handed down on 16 July 2012 (‘the Strike Out Judgment’). Following that judgment, VF successfully applied to add a further alternative measure of damages, namely a claim for damages caused by the accelerated entry of the Defendants' nets onto the market as a result of the misuse of VF’s trade secrets. The Defendants appealed against the Strike Out Judgment. In the appeal, the Defendants did not seek to strike out that part of the claim that related to the sales of the First Formula nets and they also did not challenge the new alternative claim based on accelerated entry in respect of sales of the Later Formula nets. Before the Court of Appeal, the Defendants argued that VF’s claim in relation to ‘derived products’ could only be on the basis of a head start. The Court of Appeal upheld the dismissal of the strike out application in a judgment handed down on 25 April 2013.

Below is a table showing the principal judgments so far in these proceedings and showing how I refer to them.

Other proceedings

Mr Larsen’s contract of employment with VF included a non-competition clause under which he was prohibited from involving himself in a business competing with VF for a period of 12 months from the date of his resignation from VF. Mr Larsen’s employment with VF terminated on 31 August 2004 and he was therefore bound by the covenant until 31 August 2005. In June 2005, VF applied for an interlocutory injunction at the court in Esbjerg, Denmark seeking to enforce the covenant. On 5 August 2005, the Danish court granted a preliminary injunction stating that Mr Larsen could not directly or indirectly take up employment with Danish Intection, including acting as a paid or unpaid consultant. The full case was heard in June 2006 and on 22 February 2007, the Danish court held that Mr Larsen had committed a specific act of competition by working for Danish Intection in breach of the non-competition clause. It ordered Mr Larsen to pay about DKK 400,000 in damages.

There have also been proceedings between the parties in France. In March 2007 VF obtained a search and seizure order against IIC and Dr Skovmand from a court in Montpellier. These have some tangential relevance because some of the documents in those proceedings have only become available for use in these proceedings since the Main Judgment was handed down. Those documents have helped fill in some of the gaps in understanding how Netprotect and the Later Formula were developed.

The witnesses at trial

The Claimants’ witnesses of fact were:

The Defendants criticised all the Claimants’ witnesses other than Mr Schornoz. I do not accept that criticism. I find that the witnesses were doing their best to assist the court in their written and oral evidence.

The Claimants’ technical expert witness was Mr Howe, who had given evidence at the Liability Hearing before Arnold J. The Judge commented on the scope of Mr Howe’s expertise at paragraphs 88 onwards of the Liability Judgment. At the Inquiry Hearing, Mr Howe’s evidence mainly comprised his analysis of the documents disclosed by the Defendants and the conclusions he drew from those documents. In the event, his evidence was less relevant in the Inquiry Hearing because there was less dispute between the parties about how Dr Skovmand had arrived at the Later Formula.

The financial expert engaged by VF was Mr Andrew Grantham who is a partner of AlixPartners UK LLP. He is a Fellow of the Institute of Chartered Accountants. I found much of his work clear and helpful and, as is apparent later, I have relied on much of it in coming to my conclusions.

The main witnesses for the Defendants were Dr Skovmand, Mr Larsen and Mrs Sig. In the Liability Judgment Arnold J concluded that he could not accept the evidence that Dr Skovmand and Mr Larsen gave at the trial before him. As regards both men, the Judge said that he was driven to the conclusion that they were not truthful witnesses, in particular with regard to the development of Netprotect: see paragraphs 96 and 97 of the Liability Judgment. He highlighted contemporaneous documents in which they had made statements to others that were untrue. He also referred to the deceptive and dishonest way they had, with Mrs Sig, gone about setting up the rival company and then trying to cover their tracks not only from VF but from the court. He also found that Dr Skovmand and Mr Larsen had fabricated documents. He accepted that Mrs Sig was a candid witness as regards her evidence about the Defendants’ commercial position though he thought that she knew more about the technical side of the business than she admitted.

The impression I have formed of the three witnesses is that they regard themselves as justified in taking almost any step to defeat VF’s claims in these proceedings. I do not regard it as safe to rely on much of their evidence, although not for the same reasons as Arnold J. Each of them accepts that they have little recollection regarding the events which form the basis of this claim. They accept that their evidence is largely a reconstruction of their recollection from the contemporaneous documents. However, there were two features of their evidence which concerned me. First, they are prepared to give very firm evidence as to a particular fact which they later had to accept was untrue or mistaken when they were shown a document which contradicted the conclusion they arrived at. For example, Dr Skovmand was initially very sure that references in the record of results from the field trials in October 2004 and January 2005 to tests on samples referred to as ‘mesh 156’ were irrelevant to the issues in this case because a sample described as ‘mesh 156’ must, he said, have been made of polyester rather than polyethylene. However he had to accept when challenged with other documents that these were indeed samples of PE net. Mrs Sig’s evidence about the volumes of nets that had been returned to Bestnet by Syngenta was also initially that almost all the nets had been returned by them and refunded by Bestnet. She later had to correct her evidence to make clear that in fact a large number of nets had not been returned and had instead been sold by Syngenta. Mr Larsen’s evidence about the sales of nets in the early months of Bestnet’s business was also later demonstrated to be unreliable.

I consider that they have each approached the task of reconstructing the events by starting with the end result they want to achieve and arguing back from that. This leads to them placing great weight on any indications in the contemporaneous documents that support the end result they are aiming for and ignoring or dismissing as mistaken other indications in the documents to the contrary. It is therefore better to form my own view than to rely on the rather selective memories of these three witnesses or accept their partial analysis of the documents.

So far as expert evidence was concerned, the Defendants instructed Dr Christine Meyer as their financial expert. Dr Meyer is an economist and Vice President of the well-known firm of consultant economists, NERA. Unfortunately the usefulness of her work was much diminished by the fact that she appears to have been given the wrong instructions as to the exercise she ought to be undertaking in respect of important aspects of her work. I have therefore generally preferred Mr Grantham’s approach to the issues to Dr Meyer’s.

It is useful to divide the quantification into two categories: compensation for the Defendants’ sales of nets made using the First Formula and compensation for sales of nets made using the Later Formula.

How many First Formula nets were sold by the Defendants?

Following the Liability Judgment, Arnold J ordered that the Defendants provide a witness statement verified by a statement of truth setting out the number of LLINs manufactured and sold up to 26 June 2009 (the date of the handing down of the Remedies Judgment) together with information about the profits made on those sales. The order stipulated that disclosure should distinguish between the Netprotect nets launched in October 2005 (that is the First Formula nets); the product which was submitted for WHOPES evaluation and any other PE nets made and sold by Bestnet. The evidence provided in response to this order said that only 11,500 nets made to the First Formula had been sold by Bestnet.

VF were not satisfied with the evidence provided and made an application for further disclosure. An order was made by Lewison J (as he then was) on 23 October 2009 directing that more details should be given of a particular order placed by a customer known as IDA. He also ordered that the Defendants verify information that had been set out in solicitors’ correspondence. In response to this order, Mr Larsen swore his 16th witness statement in these proceedings.

At the Inquiry Hearing before me, evidence about the numbers of nets sold by Bestnet over the period 30 January 2006 to 13 June 2013 was given in part by Mrs Sig. She produced a spreadsheet which was served on 29 January 2014 pursuant to a further court order. I will refer to this document as ‘Mrs Sig’s spreadsheet’ although she was not responsible for all the information included on it. The spreadsheet has 18 columns including in respect of every sale of nets listed, information showing the invoice date and number, the name of the customer, the quantity, size, mesh size and colour of the nets sold, the recipe used, the unit and total price. There is a column headed ‘evidence of VF bidding’ and a final column containing Mrs Sig’s comments on the transaction. These last two mentioned columns are relevant to the issue whether VF would have made these sales if Bestnet had not been in the market. It was Mr Larsen rather than Mrs Sig who had been responsible for filling in the column indicating the recipe used to make the nets in each order recorded.

It became apparent during the course of the Inquiry Hearing that Mrs Sig’s spreadsheet was inaccurate in important respects and that the Defendants’ evidence up to that point had omitted a substantial number of sales of First Formula nets. The first problem with the spreadsheet was that when completing the column which was supposed to show the recipe used for the nets included in a particular sale, Mr Larsen and his colleagues had simply assumed that the nets sold were made to whatever recipe was current at the date that the invoice was sent to the customer. Thus, because nets using the WHOPES I version of the Later Formula started to be made in March 2006 and nets using the WHOPES II version of the Later Formula started to be made in March 2007, it was assumed that any invoice bearing a date between 1 March 2006 and 1 March 2007 reflected a sale of nets using the WHOPES I version of the Later Formula.

Mr Larsen had to accept in cross-examination that any such assumption was not justified. Often nets that had been produced to an earlier recipe and kept in stock were later sold, even if production had by that time moved on to using a different recipe. It was clear therefore that some of the nets shown on the spreadsheet as being to the Later Formula recipe were in fact nets made using the First Formula. Mr Larsen acknowledged this error in his 19th witness statement made on 30 June 2014 during the course of the Inquiry Hearing. He expressed his regret about the misleading impression that had been presented by Ms Sig’s spreadsheet.

I set out in the Annex to this judgment my detailed findings in relation to the disputes between the parties over particular orders of nets. In summary my findings are as follows.

I therefore find that a total of 248,152 nets were made to the First Formula and sold by the Defendants.

What is the measure of damages in respect of First Formula nets?

The principles to be applied in compensating VF for the sales of the First Formula nets are relatively straightforward. Arnold J held in the Liability Judgment that these nets comprised a misuse of VF’s confidential information because the First Formula was based on the confidential information in VF’s Fence database. The measure of damages to compensate an injured party for an economic tort was stated by Lord Wilberforce in General Tire & Rubber Company v Firestone Tyre & Rubber Company [1975] 1 WLR 819 (‘General Tire’) at page 824 to be:

“that sum of money which will put the injured party in the same position as he would have been if he had not sustained the wrong.”

Lord Wilberforce held that in cases where the claimant and defendant are in competition with one another, making and selling competing products, the damages will include profits which the claimant can show that it has lost by reason of the wrong which has been committed. Thus, the claimant will normally receive lost profits on sales by the defendant to the extent that he can show that he would have made more sales if the defendant had not infringed. Further, in relation to infringing sales where the claimant cannot show that the defendant’s sale has caused him to lose the profit on a corresponding sale, the claimant may nevertheless recover a reasonable royalty. This principle was explained in the House of Lords’ decision in Watson, Laidlaw & Co. Ltd v Potts, Cassels & Williamson (1914) 31 RPC 104 at 119-120 where Lord Shaw described the ‘principle of price or hire’, namely that wherever an invasion of property has occurred then there should be a recompense:

“If A, being a liveryman, keeps his horse standing idle in the stable, and B, against his wish and without his knowledge, rides or drives it out, it is no answer to A for B to say: ‘Against what loss do you want to be restored? I restore the horse. There is no loss. The horse is none the worse; it is better for the exercise.’”

It was conceded by the Defendants in this case that the appropriate measure of damages for the sales of First Formula nets was the General Tire measure. VF are entitled to recover their lost profit on sales of First Formula nets to the extent that those sales replaced sales that VF would otherwise have made in the absence of Bestnet. They are entitled to a royalty on the remaining sales of Netprotect First Formula nets.

How many of the 248,152 First Formula nets would have been sold by VF in the absence of Bestnet?

The parties put forward two rival approaches as to how to work out how many of the First Formula Netprotect sales VF would have made. Mr Grantham’s reports for VF did not distinguish between the First Formula nets and the Later Formula nets in his calculations of lost profits. He approached the task on the basis that it is impossible to look at each individual sale and work out whether VF would have made that particular sale in the absence of Bestnet. The more sensible way is to assume that the other suppliers on the market, including VF, would have won Bestnet’s business in accordance with the shares in which they won the other business in the market. Thus, one should work out what VF’s share was of total sales excluding sales of Netprotect and assume that VF would have won that share of Bestnet’s sales. This method is known as ‘equi-proportional diversion’.

In applying this method, Mr Grantham assumes that the size of the total market for sales of LLINs would not have been affected by the absence of Bestnet. I accept that this is a reasonable assumption to make, given that the size of the market is overwhelmingly driven by the funds available from Governments and NGOs for buying nets and is not dependent on the numbers of suppliers. Mr Grantham also assumes (in Bestnet’s favour) that the entry of Bestnet onto the market did not result in a lowering of the market price for nets so that VF’s nets would have been sold at the same prices as they were in fact sold, in the absence of Bestnet. VF does not therefore claim a loss of profit on the sales of nets it did in fact make after Bestnet entered the market.

The Defendants accept that, if it is necessary to work out VF’s likely share of sales of Later Formula nets, then the equi-proportional diversion method is the correct one. However, they say that this approach is not justified in respect of these early sales of First Formula nets. These early sales were one-off contracts made with customers who were prepared to buy non-WHOPES approved nets. There were more sellers in that sector of the market than in the sector for sales of WHOPES-approved nets. The market shares in the subsequent years, after Netprotect achieved WHOPES interim approval are not a reliable indication of the likely substitute supplier for these early sales. The Defendants advocated an approach based on the following steps:

In my judgment the Defendants’ approach is wrong in principle as well as impossible to apply on the evidence before me. The Defendants’ submissions were in part based on their criticism of the data on which Mr Grantham based his market share calculations. Mr Grantham took his figures from Net Mapping Reports published by the United Nations Special Envoy for Malaria. The Net Mapping Reports are based on quarterly returns made by the manufacturers reporting the number of LLINs delivered by them in each territory. Mr Grantham noted in his evidence that prior to 2009, the Net Mapping Reports only recorded information on the number of LLINs supplied to Sub-Saharan Africa, not worldwide. However, his understanding is that prior to 2009, sales to Sub-Saharan Africa represented about 95 per cent of the total global market for LLINs. Mr Østergaard’s evidence for VF was that there were no significant differences in terms of VF’s competitors or their ability to supply nets between customers in sub-Saharan Africa and customers in the rest of the world. There was no evidence to contradict this. I consider that the Net Mapping Report information is sufficiently comprehensive to form an appropriate basis for a market share analysis for the early years of First Formula net sales.

Secondly, it was accepted by the Defendants that customers who bought non-WHOPES approved nets from Bestnet in the early months could equally well have bought WHOPES approved nets. There was no evidence of a price differential between approved and non-approved nets. I accept Mr Østergaard’s evidence that over the period when Bestnet was making its sales to these customers, VF was also making sales to them or at least being invited to discuss supplies to them. There does not seem to be any reason why the people who bought First Formula nets from Bestnet would not have considered VF’s PermaNet product if Bestnet had not been available. There is no reason why VF should not have got its market share of that business.

Thirdly, the evidence provided by Mrs Sig about VF’s chances of winning the individual sales in the absence of Bestnet on the market seems to me very flawed. This evidence was summarised in the final comments column on Mrs Sig’s spreadsheet. Much of it is based on comments that customer representatives made to her in her discussions with them. These seem to me a flimsy basis on which to make a finding of fact. Of course, representatives will make encouraging noises to a new supplier and join in disparaging the opposition when they meet to discuss sales. That does not mean they would not buy from VF if Bestnet were not present on the market. Mrs Sig’s evidence also seemed in part to be based on an assumption that customers at that time distinguished between polyester and polyethylene nets and bought particular proportions of each. She therefore concludes that customers who bought PE nets from Bestnet and PES nets from a supplier other than VF would not have bought PES nets from VF in the absence of Bestnet. Although there is some debate in WHO circles as to the relative merits of polyethylene and polyester nets, the evidence shows that at the time of these sales, the two kinds of nets were substitutable in the eyes of customers and may well still be. On balance I prefer Mr Østergaard’s evidence that it is only a small part of the market that is tendered specifying PE or PES nets. In the absence of Bestnet, a customer may well have bought PES nets from both VF and its existing PES net supplier.

In my judgment, therefore the best approach is to apply the equi-proportional diversion approach to these early sales of First Formula nets. I therefore endorse the approach that Mr Grantham has taken.

The figure he arrived for VF’s share of overall LLIN sales (excluding Bestnet’s sales) in 2006 was 73.9 per cent.

The next step in the calculation is to apply that market share figure to the sales of First Formula nets made by Bestnet as I have found earlier in this judgment. I have found that the total number of First Formula nets sold by Bestnet was 248,152. The calculation of the General Tire measure of damages should therefore be based on the assumption that 183,384 nets sold by Bestnet would have been sold by VF.

In respect of those lost sales of First Formula nets, what is the lost profit for which VF is entitled to be compensated?

Mr Grantham went on to work out how much incremental profit VF would have earned on each of the additional nets it would have sold if Bestnet had not made any sales. He computed these using the management accounts of VF Holdings. He calculated the revenue per net and the costs of sales. The incremental profit per net in the year 2006 is US$2.02 per net.

The Defendants accept that these profit figures have been correctly computed.

The final step is therefore to apply that profit per net to the number of First Formula nets sold by Bestnet that would, applying the equi-proportional diversion method, have been sold instead by VF. This produces a figure of US$370,436 lost profit.

Two points need to be considered at this stage. In his computation, Mr Grantham deducts the costs that he says VF would have had to incur to buy two additional stenter machines in order to make the additional nets. My understanding however is that this deduction relates to the later much more substantial sales of Later Formula nets that Mr Grantham includes in his calculation. VF would not have had to buy more stenters to produce the relatively modest numbers of additional nets it would have sold in this period.

The second point is whether there should be a reduction in the claim for loss of profit because some of the benefit of the profit which would have been made on the nets would not have been enjoyed by the VF group but rather by the holders of minority interests in the group company which makes the sales of nets, that is Vestergaard Frandsen SA. Evidence about the corporate structure of the VF group was given by Nicolas Schornoz who is currently the Chief Financial Officer of that group. He exhibited the Consolidated Financial Statements of the group from April 2005 until the year ended 30 June 2006 which shows income before minority interests of US$24,659,000; minority interests of US$3,103,000 and hence net income after minority interests of US$21,556,000. Thus it appears that over the course of the period about 12.6 per cent of Vestergaard Frandsen SA was not owned by the group but by minority interests. VF accepted that its loss of profit claim for that year should be reduced by the same percentage. This follows from the Court of Appeal’s judgment in Gerber Garment Technology Inc v Lectra Systems Ltd and another [1997] R.P.C. 443.

Having made that reduction, I calculate that the value of VF’s loss of profits claim in respect of First Formula nets is US$323,761.

The royalty for the remaining sales of nets

The second part of the General Tire measure of damages is to work out what royalty VF should recover in respect of sales of Netprotect nets that would not have been replaced by sales of VF nets in the absence of Bestnet on the market. That amounts to 64,768 nets (that is the remaining 26.1 per cent of 248,152). The royalty would, under this hypothesis, be payment for permission to use the confidential information that Arnold J held had been used by Dr Skovmand to arrive at the First Formula.

The principles to be applied in assessing were set out by Arnold J in Force India Formula One Team Limited v 1 Malaysia Racing Team SDN BHD [2012] EWHC 616 (Ch); [2012] R.P.C 29. There was claim for loss of profits in that case, the damages were computed entirely on the value of the unlawful use that had been made of the claimants’ confidential information. The primary basis for the assessment was, Arnold J held, to consider what sum would have been arrived at in negotiations between the parties, had each been making reasonable use of their respective bargaining positions, bearing in mind the information available to the parties and the commercial context at the time that notional negotiation should have taken place. The principles governing that hypothetical negotiation were that:

Further light was shed on this topic by Newey J in 32RED plc v WHG (International) Limited & ors [2013] EWHC 815 (Ch). In that case the defendant was found to have infringed the claimant’s trade mark in the name ‘32Red’ by operating an online casino using the brand name ‘32Vegas’. Two specific issues arose as to the value of the hypothetical licence to use that name:

As to the first question, Newey J held that the parties are taken to have been willing to make a deal even if one or both of them would not in reality have been prepared to do so. It is also assumed that the parties would have acted reasonably regardless of whether that would in fact have been the case. Thus, particular character traits of the parties should be disregarded, for example whether they are easy-going or aggressive. The hypothetical negotiation is designed to establish the value of the wrongful use to the defendant and not some objective figure as between hypothetical persons negotiating for a hypothetical licence. The negotiation would be one between the actual parties, albeit that they are to be treated as parties willing to deal with each other with a view to reaching a reasonable result: see paragraph 32 of his judgment.

As to the second question, it is important to distinguish between two different principles. The first is the principle that a defendant cannot defeat a claim for infringement by arguing that he could have achieved the same result without infringing the claimant’s rights. This is the principle established in The United Horse Shoe and Nail Company Ltd v John Stewart & Co (1988) LR 13 App Cas 401 where the pursuer’s patent for a machine for making horse-shoe nails had been infringed. The defenders argued that they could have made similar nails without infringing the patent and that the pursuer had therefore suffered no loss. Lord Macnaghten held that such an argument ‘appears to be beside the mark’. However, as both Newey J in 32Red plc and Arnold J in Force India observed, Lord Macnaghten’s comments were made in relation to a claim for loss of profits, within the first category of damages posited by Lord Wilberforce in General Tire. Newey J held that it is right to take into account, when considering the respective hypothetical negotiating positions of the parties, their awareness of what alternatives routes to the same end were available to the defendant if he did not obtain a licence of the confidential information. This approach was also adopted by the Court of Appeal in Force India Formula One Team Limited v Aerolab SRL [2013] EWCA Civ 780, [2013] R.P.C. 36 where Lewison J said:

“107 … In any negotiation the parties to the negotiation will be considering what their alternatives are to doing the deal. There is no reason why a hypothetical negotiation should be any different in that respect. It is, of course, different from a real negotiation in one respect because in the hypothetical negotiation not doing the deal at all is not an alternative. … ”

In 32Red Newey J therefore took into account the fact that the defendant in that case could have rebranded the online casino with a non-infringing name and that would have been a factor influencing the outcome of any hypothetical negotiation. He held that the starting point when considering that negotiation was the economic benefits that the parties would have expected from use of the mark but that “the possibility of re-branding would have been of central importance in the hypothetical negotiation”. Further he said:

“…32Red, as a hypothetical willing licensor, is to be taken, I think, to have recognised that it could not insist on being paid a sum out of proportion to the financial advantages that the defendants stood to obtain by using the name 32Vegas rather than re-branding at once. …”

On the other hand, he recognised the benefits of keeping the name 32Vegas once the defendant had set up its business using that name but been informed of the claimant’s mark. They would suffer the cost and inconvenience of rebranding and risk disrupting their business.

In the present case, the parties’ approaches to the calculation of the royalty were very different. Dr Meyer for the Defendants bases her work on the assumption that Bestnet would have been able to launch and sell its Netprotect nets even if it had not misused the confidential information. Dr Skovmand would, she assumes, have devised a series of trials and then undertaken testing and development of the product. The additional costs would have been within the range of US$7,500 and US$27,303 (including for example trips to India and costs of bioassays). She notes that it is reasonable to assume that there would have been some delay in its product launch but that there have been no findings so far as to how much delay. She therefore calculates loss of profit on the basis of four scenarios ranging from no delay resulting from the additional tests to a delay in the launch date of the First Formula nets product of six months, taking the launch to April 2006.

So far as the royalty rate is concerned, she bases her calculation on the ‘anticipated economic benefit’ of the Fence database to the Defendants. She says:

“54.

Properly constructed, the hypothetical negotiation reflects the relevant expectations and market factors that would have affected a real world licensing negotiation at the time of the negotiation. For example, the construction of the hypothetical negotiation takes into account the expectations of the negotiating parties regarding future sales of and profits from the products that were sold wrongfully, the expected costs imposed on the Claimants that flow from the license, and the availability of and risks associated with reasonable, potential alternatives, if any. In addition, licensing experience in the same or related products or, more generally, by either negotiating party, may be relevant to the analysis of the hypothetical negotiation.”

She determines first the outer ranges of the royalty, that is the highest royalty that Bestnet could pay and still be better off taking a licence and the lowest royalty that VF could charge and still be better off granting a licence. As to Bestnet’s position, she says that Bestnet would not have been willing to pay more in royalties than the sum of the forgone benefit due to later entry into the market and the costs Dr Skovmand would have incurred in re-creating the Fence database. This then ties in with her conclusions on delay, because clearly the value of the royalty, if the delay in coming up with a legitimate recipe was only 8 weeks, should be set at a level that reflects 8 weeks’ worth of sales. She works out how many sales would have been made in that 8 week period by looking at Mrs Sig’s spreadsheet to see how many sales were in fact made during that time. If there were sales made in that period, she regards those sales as being ‘at risk’ from the delay that would occur if no licence was obtained and she works out the profit associated with those ‘at risk’ sales to arrive at a discounted royalty negotiated as at October 2004. This would reflect the fact that with the licence, Bestnet would be able legitimately to make those sales and without the licence, it might not. She also takes into account the savings in costs as a pointer to the royalty that Bestnet would have been willing to pay.

Next she considers the lowest sum that VF would be willing to accept and still be better off by granting a licence. This is based on the assumption that VF is not losing any of the sales that Bestnet would make because this royalty is only paid on the third class of General Tire sales. Thus any royalty is better than nothing for VF. She notes that the parties to this hypothetical negotiation would have equal bargaining power and that the bargain struck would therefore fall at the mid-point of the bargaining range. She therefore calculates the reasonable royalty as a lump sum equivalent to half the profit or costs saving that Bestnet would have earned, discounting that sum to reflect the fact that it is paid in October 2004.

Mr Grantham’s approach is first to set out some comparable royalty rates in licensing agreements entered into by the Defendants, other licensing agreements recorded in a database the RoyaltySource Intellectual Property Database and published studies of royalty rates. He concludes that none of these sources is particularly helpful beyond showing that royalty rates are in a range from 3 per cent to 8 per cent. He then considers the profit expectations of VF and the Defendants. He has been provided with the Defendants’ business plan prepared in 2004. This shows that substantial sales of nets were expected but that these would mainly be polyester nets rather than PE nets. That plan, and the actual financial results he has analysed in his report, lead him to believe that the financial results might not in this case be a useful guide to the Defendants’ profit expectations as at the date of the hypothetical negotiation. He therefore turns to considering VF’s financial information on the basis that VF would have given greater weight to the actual profit levels they were achieving rather than the untested forecast information that would have been available to the Defendants. Looking at VF’s incremental profit on their own sales, a 3 per cent royalty on the Defendants’ average sale price would amount to 6 per cent of the profit (on the basis of 2008 figures, being the earliest year Mr Grantham includes in his table) and a royalty of 8 per cent would amount to 15 per cent of profits. He concludes that a range of 3 to 8 per cent is the appropriate range to be used.

I have not found either Dr Meyer’s or Mr Grantham’s reports particularly helpful in arriving at a conclusion on the appropriate royalty rate. Both seem to me to be flawed. Dr Meyer’s approach is problematic because it seems to me most unlikely that Bestnet would have agreed to pay an upfront lump sum for a royalty. At the time of the hypothetical negotiation, there was no successful formula for bed nets in existence and the parties did not know whether use of VF’s information by the Defendants would ultimately lead them to selling any nets at all. No one knew how much more work would be needed to build on the information in the Fence database to produce a saleable net. We are considering here only the First Formula nets. We do not know whether those nets would have passed WHOPES evaluation because they were never sent for that evaluation. Bestnet would be taking a licence of the information knowing that they would have to invest in more bioassays and other testing of samples and might ultimately fail to produce a net capable of competing with the existing products on the market.

On the other hand, it seems equally unlikely that VF would have been prepared to licence the result of the Fence database for a very low, one-off, lump sum to reflect the uncertainty surrounding the value of the information. They knew from their own business that if an effective net can be produced, the volumes of potential sales are very large indeed and the profit margin on those sales also appears to be very generous.

Dr Meyer’s method of calculation gives undue prominence to the parties’ consideration of how long it might take Dr Skovmand to come up with the First Formula without using the confidential information. I agree that this is an element to be taken into account and that VF would have had regard to the fact that Dr Skovmand, as a renowned expert in the field and with many years’ experience in this kind of work, might well be able to come up with his own recipe for bed nets without VF’s help. But this is not the only factor to consider. Dr Meyer’s computation of the maximum royalty that Bestnet would be prepared to pay, and hence her conclusion as to the likely outcome of the negotiation, is entirely based on calculating how long it would have taken Dr Skovmand to come up with the First Formula by his own efforts and how many of the actual sales made during that period would have been ‘at risk’ if a corresponding delay occurred. Some degree of hindsight is permissible in this exercise but Dr Meyer’s method entirely excludes an important element that would have influenced the negotiations, namely the parties’ uncertainty over whether the Fence database information would ultimately prove to be of no value; whether it was a goose about to lay a golden egg with very little additional effort, or was something in between.

Mr Grantham’s approach arrives at figures which cover a broad range. He is also influenced towards a higher figure by Mr Rohde’s evidence, which Mr Rohde emphasised when being cross-examined, that VF would not in fact have been prepared to grant a licence to Bestnet or anyone else to make nets using their confidential information because there would be no benefit to them in assisting a competitor to enter the market with a rival net. I do not regard this as a helpful consideration for two reasons. First, as Dr Meyer pointed out, we are considering here a royalty on the sales which it is assumed are not sales diverted from VF’s product and which VF would not have made in the absence of Netprotect on the market. Secondly, the case law makes clear that the hypothetical licensor cannot be heard to say that he would have refused to grant a licence at all – that is not regarded as a relevant factor. If one increases the licence fee on the grounds that the licensor would be reluctant to grant a licence, that appears to be reintroducing that element of unwillingness by the back door.

I have concluded that the correct royalty rate is a per unit rate at the lower end of the range put forward by Mr Grantham, that is 4 per cent. That seems to me to reflect the value to both parties of VF’s contribution to what might turn out to be substantial future sales of an effective competing net, whilst not committing Bestnet to paying for information that might not lead to any sales at all.

There is a difficulty in coming to the final figure for this royalty payment. Although I have determined the number of First Formula nets sold, it is not clear at what price they have been sold. There is no documentation surrounding the sales of the Karur Stock Nets. Mr Grantham’s report gives an average LLIN sales price for the Defendants’ nets for the year 2006 as US$4.73. But that may cover a mixture of First Formula and Later Formula net sales since his figures do not distinguish between them.

The total sales value of lines 1 – 15 on Mrs Sig’s spreadsheet shows a total value of US$846,536 earned on the sale of 186,352 nets. Assuming that 26.1 per cent of those sales are to be subject to a royalty, that leads to proceeds of sale of US$220,946 on the sale of 48,638 nets or an average price of US$4.54 per net. A royalty of 4 per cent gives a per net figure of US$0.18 royalty.

The total number of nets subject to the royalty is 26.1 per cent of the nets from lines 1 – 15 of Mrs Sig’s spreadsheet plus 26.1 per cent of the Karur Stock Nets, being 16,130 further nets. That makes 64,768 nets in all, subject to a royalty of 18 cents making US$11,658 royalty due on sales of First Formula nets in addition to the lost profit.

There is no reason to reduce the royalty to reflect minority interests in Vestergaard Frandsen SA because any licence would have been granted by the company at the head of the group.

Conclusion on damages for sales of First Formula nets

I therefore award damages of US$335,419 for the sales of First Formula nets.

The correct measure of damages

One of the main areas of dispute at the Inquiry Hearing was the principles that should be applied when considering the damages to which VF are entitled in respect of sales of Later Formula Netprotect nets. VF’s primary case is that there is no justification for distinguishing between the First Formula and Later Formula nets – the General Tire measure of damages should apply to all Bestnet’s sales up to September 2013. In the Re-Amended Points of Claim VF put their claim in three ways:

In support of their primary contention, VF argue that this is the result of applying the general law on remoteness and causation of loss. The damage that VF have suffered as a result of Bestnet’s sales is precisely the kind of damage from which the laws prohibiting unauthorised use of confidential information are designed to protect them. These damages are not therefore too remote from the wrong committed. Further, they argue that there is a strong causal link between the misuse of the confidential information and the sales of all the nets. That is so even though Arnold J held that the Later Formula nets did not of themselves comprise misuse of confidential information but were derived from that misuse. There is nothing here, they argue, that constitutes a break in the chain of causation between the wrong and the damage. They acknowledge that if General Tire damages are payable on all sales of Netprotect, then their claim is worth a great deal of money. But this results, they say, not from exaggerating their claim but from the fact that Bestnet sold 57 million nets using recipes which either constituted a misuse of VF’s confidential information or which derived from that misuse.

The Defendants’ case in respect of damages payable for Later Formula nets evolved somewhat during the course of these proceedings. By the time of closing submissions, Mr Wilson QC for the Defendants argued that the only proper approach to damages is to identify such misuse of the confidential information as had occurred and assess the damages properly to be awarded in relation to that misuse. He put forward a two-pronged approach to the quantification of damages for Later Formula nets. First, the Defendants acknowledge that there was some use by Dr Skovmand of VF’s confidential information in arriving at the Later Formula and in obtaining the WHOPES interim approval. They accept that VF is entitled to be compensated for that use. They propose a modest consultancy-fee type award of damages in respect of that. They submit that US$ 10,000 ‘would generously meet the case’.

Secondly, they accept that in so far as the misuse of VF’s confidential information accelerated Bestnet’s entry into the market, then VF would be entitled to damages in respect of sales that were made by Bestnet in the period when they would not have been on the market but for the use of the information. The Defendants further accept that any such sales should be compensated in accordance with the principles set out in General Tire. However, the Defendants say when one looks at the facts here, it is clear that Dr Skovmand could have come up with a net without misuse of confidential information quite quickly and certainly in time for their application to be considered at the same December 2007 WHOPES meeting as the application in respect of the Later Formula nets was in fact considered. They would still have got their WHOPES approval in March 2008. There was therefore no accelerated entry in respect of Later Formula nets.

Neither side was able to point me to an authority in which the proper measure of compensation for derived products has been determined. In the Remedies Judgment, Arnold J noted that “[t]here is curiously little authority” on the issue, though he was focusing there on the grant of injunctive relief to restrain sales of such products. It was the novelty of the point that subsequently dissuaded Arnold J and the Court of Appeal from striking out VF’s claim for damages in relation to the Later Formula nets.

I should say first that I reject the Defendants’ submission that Arnold J’s refusal in the Remedies Judgment to grant an injunction to stop sales of Later Formula nets means that I should not award any damages for those sales. That refusal was not based on the assumption that there was no cause of action at all in relation to those sales. So far as the comments of the Court of Appeal in the Liability CA Judgment are concerned, Jacob LJ referred to Dr Skovmand having taken a short-cut by using the Fence database information and noted that it was a matter of debate how much time and trouble was thereby saved: see paragraphs 24 and 57 of the judgment. But, again, the Court was not dealing with the question of the proper measure of damages.

More relevant are the comments of Arnold J and the Court of Appeal when considering the Defendants’ application to strike out those parts of VF’s Amended Particulars of Claim which sought damages for sales of Later Formula nets. Arnold J stated that he did not find the question of law raised by the application an easy one. He regarded it as surprising that VF did not draw a distinction between the First Formula and Later Formula nets, given the findings in the Liability and Remedies Judgments. But he could not say at that stage that VF's claim to lost profits or a notional royalty in respect of the Later Formula nets was unsustainable.

In the Strike Out CA Judgment, Floyd LJ referred to the judgment of Laddie J in Ocular Sciences Limited v Aspect Vision Care Ltd (Part 2) [1997] RPC 395 as a case where a distinction was drawn in the context of the grant of injunctive relief between sales of products which made use of the confidential information and sales of derived products. Laddie J referred to derived products as being products which:

“… do not themselves continue to incorporate or disclose confidential information although they were brought into existence or were perfected or owe their commercial success to the fact that confidential information was used in the past…” (page 396)

Floyd LJ noted VF’s submissions that the factual issue as to how much use Bestnet had continued to make of VF’s confidential information in making the derived products remained to be elucidated. He referred to the cases of Dowson & Mason v Potter and anr [1986] 1 WLR 1419 and Saltman Engineering v Campbell Engineering (1948) 65 RPC 203. He concluded that there was no authority directly on point as to the assessment of damages in a breach of confidence case involving a derived product although it was clear that the underlying principles must be those expressed by Lord Wilberforce in General Tire and by Lord Shaw in Watson, Laidlaw. Further, he said this:

“29.

… I do not consider that the distinction which the Judge drew in the remedies judgment for the purposes of deciding to grant an injunction necessarily means that the only basis for assessment of damages in the case of the derived products will be the head start or accelerated entry basis. Injunctions and damages are distinct remedies and the principles which govern their availability are not the same. …

30.

Still further, it seems to me that, when it comes to considering damages, the distinction between the two classes of product sold by Bestnet may not be as material as it is when considering the grant of an injunction. Both classes, to some, although a differing degree, benefit from the VF confidential information. Whether it is right, in the end, to limit VF to head start damages in respect of the derived products is a decision which can only properly be made when the extent of that benefit has been established on the facts. …”

The parties at the Inquiry Hearing drew my attention to two further authorities, though they recognised that neither was particularly helpful. In Gerber Garment Technology Inc v Lectra Systems Ltd and another [1995] R.P.C. 383 the issue was whether the claimant could claim for the profit lost on ancillary supplies such as spare parts and maintenance servicing in respect of sales of fabric cutting machines that were made using confidential information. It was held that since the claimant would have earned that profit if the customers had bought its machines rather than the defendant’s infringing machines, the additional lost profits were recoverable. There, however, it was not a question of the extent to which the spare parts or services also used the confidential information but rather that the reduction in machine sales suffered by the claimant led to lost profit not only on those sales but of the ancillary supplies as well. In the present case, it is not suggested that sales of the Later Formula nets were in any way ancillary to the sales of the First Formula nets. In Work Model Enterprises Ltd v Ecosystem Ltd and Clix Interiors Ltd [1996] FSR 356 the claimant sought damages in respect of sales of the defendant’s product on the grounds that the brochures used by the defendant to promote its product infringed the claimant’s copyright although the products themselves did not infringe. The court struck out the paragraphs in the claimant’s pleading on the grounds that there was no sufficient nexus between the sales of goods and the infringing wording of the brochure. In that case, the nexus between the brochure and the sales of products was so tenuous that the court did not have to consider precisely how close the link has to be in order for a claimant to be entitled to compensation.

I do not accept VF’s argument that, applying the basic principles of compensation for tortious conduct, results in liability to compensate for all foreseeable loss, subject only to the question of remoteness and causation. This is particularly not the case where the loss suffered is purely economic loss rather than physical injury or damage to property. For example, the authorities regarding liability for negligent advice establish that it is not enough to show that the loss was foreseeable and not too remote: see Caparo Industries Plc v. Dickman [1990] 2 AC 605 whereLord Oliver of Aylmerton acknowledged that in relation to pure economic loss, the concept of foreseeability of harm as a single test, even a prima facie test, of the existence of the duty of care must be discarded if the law is to be kept within the bounds of common sense and practicality: see page 633 of his speech. There is a broader policy apparent in the cases that for economic loss, liability for tortious conduct should be kept within reasonable bounds. The previous judgments in these proceedings show the courts’ concern that compensating VF for sales of Later Formula nets to the same extent as it is compensated for sales of First Formula nets may not be a fair reflection of the harm that VF has suffered as a result of the Defendants’ conduct. I share that concern. It is not a concern arising from the amount of money claimed. If Bestnet had discovered that the First Formula could not be bettered and had sold millions of First Formula nets over the years, its liability to VF may have been very substantial. But on the facts of this case, it would be wrong in my judgment to base the quantum of damages on an assumption that having misused the confidential information, the Defendants must compensate VF by making good their profit on lost sales and paying a royalty on other sales indefinitely, unless and until they can point to some intervening event that produces a definite break in the causal chain. In this case, as I describe shortly, there was substantial further experimentation and expertise brought to bear by Dr Skovmand in order to make the jump from the First Formula to the Later Formula. The differences between the two formulae were material. The successful formula which led to the attainment of WHOPES interim approval and which then led to the substantial volumes of sales was the result of further development beyond the work which Dr Skovmand had carried out when working for VF and which he wrongly made use of when he was working for the Defendants.

What then is the proper approach? I consider that the two-fold approach put forward by the Defendants is a fair and proportionate approach. One head of damage should be a lump sum, quasi-consultancy fee to reflect the extent to which the sales of Later Formula nets were brought about by the use of the confidential information by the Defendants. In order to decide how much this should be, one needs to consider how closely linked the Later Formula is to the First Formula and in what other ways the Defendants made use of that information, beyond simply building on the experimental results in the Fence database to arrive at the Later Formula.

The second head of damage is a payment of compensation in respect of sales made in any period when the Defendants were on the market selling Netprotect when they would not have been on the market if they had not misused VF’s confidential information.

The quasi-consultancy fee for use of the confidential information in arriving at the WHOPES approved, Later Formula Netprotect nets

How did the Defendants arrive at the Later Formula?

Following the Liability Trial more information came to light about how the Later Formula was devised. The chronology is not straightforward because each step has various stages; the production of the samples of different recipes at the Siva factory in Karur, the start of tests on samples made to that recipe to work out the regeneration time for the sample and the wash resistance of the sample and the analysis of those results as they are reported and the creation of further recipes and further samples building on those results. The development is therefore not entirely linear because whilst tests are continuing on early samples, new samples are being produced and subjected to tests as well. I must also make clear that the narrative that follows concentrates on the samples that proved to be main steps on the road to the Later Formula. It is important to bear in mind that many other recipes were made up into samples and subjected to testing but proved to be less successful. This is apparent from the much fuller description of events set out in the Liability Judgment. In addition to many different combinations of the three main additives, there were other additives that were tried and different formulations of polymer used in different samples to find out what it was that made the difference between an effective and an ineffective net.

The key question is what information Dr Skovmand had in front of him at the time when the samples of Later Formula were made in March 2006 and sent to the Montpellier Lab to be tested for WHOPES I evaluation and how much of that information relied on the testing of nets made up to the First Formula. The development of the Later Formula proceeded in the following way. Numbers in square brackets refer to paragraphs in the Liability Judgment.

Step 1: October 2004/January 2005 manufacture and testing The first samples of yarn and net were made for the Defendants at Mr Sivasamy’s factory in Karur between 25 – 27 October 2004. These samples were entered on the first lines of the Netprotect database. The samples all contained [Additives A, B and C] as well as deltamethrin. Some of the recipes used were the same as recipes taken from the Fence database (in particular [Figure 1] parts [Additive A], [Figure 2] parts [Additive B] and [Figure 5] parts [Additive C]. Other samples were made to the First Formula. Some of the samples contained double the concentration of insecticide.

Nets from this October 2004 sample production were sent out for a variety of tests. Eight pieces of First Formula net were sent by the Defendants to the Montpellier Lab for bioassay in mid-December 2004 [414]. This was referred to as sample 19. Arnold J found that the primary purpose of these tests was to test the efficacy of the prototype net according to a protocol approximating to the WHOPES Phase I protocol [415]. Some samples of First Formula net were also sent to the University of Montpellier for chemical analysis in January 2005.

In January 2005, the Montpellier Lab sent test results back to the Defendants reporting on the results of the tests after 20 washes. The results were very encouraging.

After the Montpellier Lab tests had been completed, the nets they had tested were sent to another laboratory, the Walloon Agricultural Research Centre in Gembloux, Belgium for chemical analysis of the yarn to find out how much insecticide was left inside the yarn. The results of those tests came back in April 2005. These tests showed that there was still a substantial amount of insecticide left in the First Formula yarn even after 30 washes but that the concentration of it varied across the net [440].

The Montpellier Lab’s final report on the First Formula net samples that had been sent to them in December 2004 was sent to the Defendants on 3 May 2005. This final report played an important role in subsequent events and I shall refer to it as the Montpellier First Formula Net Report. Arnold J noted that Bestnet posted a copy of the report on its website as part of the promotional materials for Netprotect. It remained on the website until after the start of the Liability Trial even though by that time, of course, the nets being sold were not First Formula but Later Formula nets [447].

Some of the nets that were made in the October 2004 sample production (including some First Formula net samples) were sent for field trials in Burkina Faso. Test results started to come in on those nets as they were progressively washed, allowed to regenerate, and washed again, and then tested to ascertain their continued toxicity for mosquitoes. Part way through the first batch of tests in Burkina Faso it became apparent that there was a problem because the masterbatch used in making the samples had been too concentrated, so that the amount of insecticide and additive was not uniform across the surface of the net. It was necessary to restart the tests. Between 18 and 25 January 2005, further sample nets were produced at the Karur factory using a lower concentration of masterbatch and using a total of 8 different recipes. These included samples made to the First Formula. These were sent out to Burkina Faso and further wash tests were carried out on them.

In July 2005 there was also a production run of nets, in addition to the creation of samples of different recipes. Arnold J held that the production run in July 2005 was made to the First Formula: [471].

Step 2: the Statistix exercise carried out in July 2005 The next important step in the development of the Later Formula was something that was not in evidence during the Liability Trial. Dr Skovmand said that in July 2005 he decided to pursue the idea of increasing the [Additive C] content of the yarn and trying different [Additive A] concentrations. He used a computer modelling programme called Statistix to work out, through linear regression analysis, which of the components in the recipes being tested appeared to have the most significant effect on the outcome. He fed into the model the results that had come through so far from the Burkina Faso bioassay tests on 16 different samples. These samples which had been made in Karur in October 2004 and January 2005 were to five different recipes, one of which was the First Formula. Some of the samples included double the dose of deltamethrin. By July 2005, the tests in Burkina Faso on the samples had progressed as far as measuring the effectiveness of the samples after 21 washes.

The results of the Statistix exercise led Dr Skovmand to conclude that the [Additive C] concentration was [REDACTED]. This meant that changing the [Additive C] concentration was more likely to affect the end result than changing the [Additive A] concentration. This prompted him to conclude that two other recipes were likely to be an improvement on what had been produced so far. These became sample 108 (made to a recipe using [Figure 1] parts [Additive A], [Figure 2] parts [Additive B] and [Figure 6] parts [Additive C]) and sample 109, also referred to as sample C (made to a recipe using [Figure 3] parts [Additive A], [Figure 2] parts [Additive B] and [Figure 6] parts [Additive C]).

Step 3: the testing of samples 108 and 109 Nets made up to these recipes [REDACTED] were produced for the Defendants by the Siva factory in Karur in July and August 2005 and sent out to Burkina Faso for more testing. In addition more samples using the First Formula were made up at the same time and sent for testing to provide a comparable reference point against which to assess the new recipes. Dr Skovmand explained why he had further samples made up to the First Formula at the same time as making samples to recipes he thought would perform better:

“A. It does not mean that the reference sample was particularly good. The point was that that LIN [the Montpellier Lab] made a test on the recipe [First Formula] and LIN is the golden laboratory, the golden standard, so therefore until I had [Later Formula] as the confirmed recipe, I used [First Formula] as a reference recipe. So, all samples were compared to that. In the way that with -- not that I took the original 100 number 19 [samples] or number 1 [samples], but I remade the same recipe together with the test samples so it was made under the same circumstances, the same HTP polymer because producer often changed and I wanted to be the same so I could compare. That was the meaning of it.

Q. I understand. So, you did not simply take the old net from way back.

A. No.

Q. Each time you wanted to use it as a reference you would remake the [First Formula], have it there and then compare them?

A. Yes.

Q. Then you would know they were made on the same day, by the same factory?

A. Exactly.

Q. With the same ----

A. Sure.

Q. That was an important way for you -- important assistance for your development of the Netprotect net?

A. Yes. What, I did not understand once again?

Q. That was an important thing for you to be able to do, to compare it to the reference sample as you developed the net?

A. Sure.”

Additional samples to different recipes were also made in July 2005 and tested at the same time but the expectation, based on the Statistix exercise, was that samples 108 and 109 would perform strongly.

Again, the July/August 2005 samples were not only sent to Burkina Faso to start field trials but were also sent to the Montpellier Lab to assess the regeneration time for the nets under conditions more closely approximating those that would be used by WHOPES. The results of the Montpellier Lab tests on samples 108 and 109 were reported to the Defendants in about November 2005. These showed that the regeneration time for samples 108 and 109 was four to five days - longer than the three days that would be acceptable for WHOPES.

Meanwhile, in Burkina Faso the second series of tests were being carried out on a range of samples including those made to the First Formula and those made to the sample 108 and 109 recipes. This new set of tests had started in September 2005 and continued throughout the rest of 2005. The results of each bioassay test were reported to Dr Skovmand. The results were not as Dr Skovmand had predicted and when he caused this to be investigated it became apparent that there was a problem with a breakdown of the mosquito colony being used for the tests. At the end of February 2006 therefore the tests were restarted. By March 2006 it became apparent from the Burkina Faso results that the recipe of sample 109 (that is [Figure 3/Figure 2/Figure 6]) was the best performing net.

Step 4: choosing the Later Formula and making sample 114 By this time, Dr Skovmand says, he had already come to the provisional view that the formula for the nets to be sent for WHOPES evaluation should be a modified version of the recipe used for sample 109. The recipe for sample 109 was, as I have stated, [Figure 3/Figure 2/Figure 6]) and had resulted, according to the Montpellier Lab tests, in a net with a longer regeneration time than was ideal. There was another problem with sample 109 namely that it was made too thin and was too fragile for use. [REDACTED] It is also more cost effective if one can achieve the same result by using less of one additive rather than more of the other. So Dr Skovmand decided that the best formula for the nets was likely to be [Later Formula]. He therefore directed that samples made to that recipe be produced.

The first sample made to a recipe containing the [Later Formula] proportions was sample 115 made in January 2006. However, it was subsequently decided that the samples to be sent to WHOPES should contain a higher dose of deltamethrin than had been incorporated into sample 115. A further sample, sample 114 was made in March 2006. It contained [Figure 7] parts deltamethrin, [Figure 3] parts of an additive of a different brand but equivalent to [Additive A], [Figure 2] parts of an additive equivalent to [Additive B] and [Figure 4] parts of [Additive C]; it also contained a small amount of [Additive M] [509].

Arnold J found that this, rather than the First Formula, was the recipe for the nets that were sent to WHOPES for evaluation in April 2006 [513]. It is noteworthy therefore that the recipe that was used for the samples sent to WHOPES was chosen before the results of tests on any samples made to that recipe had been received, either tests at the Montpellier Lab or in Burkina Faso.

Dr Skovmand’s evidence was that there was some urgency about submitting the samples for WHOPES evaluation and that is why the samples to the new Later Formula were sent off without waiting for testing. (‘LIN’ here refers to the Montpellier Lab)

“MR. PLATTS-MILLS: Dr. Skovmand, a little while ago you said that submitting for WHOPES in June 2006 was not an option. Yes? You did not want to do that; you did not want to put it off.

A. Yes.

Q. Why was that? Why was that a problem?

A. Because there was -- I was under pretty heavy pressure from Bestnet to put in something, because they had economic problems because of these court cases. You know, from start, from start, your client managed to delay Bestnet – Intection they were called at that time -- three critical months. In those three critical months, three other products came into the WHOPES process, and the WHOPES process starts at LIN. LIN is a French research organisation. They are not paid for doing this and that. So, the French government can only run a low number of these tests for WHOPES every year. You have to concentrate on the results. Therefore, if you are delayed in that queue, you are not delayed three months, you are delayed a year. So, if I delay this to June 2006, Bestnet would be left not six months, but another year, and they would be out of business before they were started. So, that was the pressure; and I did not like it, but I had to face realities.

Q. Sounds like commercial reality?

A. Yes.

Q. Just so that you understand, you realised that there was that commercial pressure at that time; yes?

A. Sure.

Q. Because of the financial issues and because you needed to get into the queue for WHOPES?

A. Yes.

Q. And if you were delayed a few months, you might end up being delayed a year or more?

A.

Yes, exactly.”

I accept that evidence since it explains the rather bold step of submitting samples made to a recipe which had not in fact undergone much testing before the application was sent.

In summary therefore at the time that the Later Formula nets were produced to be sent for WHOPES evaluation, Dr Skovmand had available to him the following information:

Step 5: the WHOPES II formula Arnold J found that the nets that were submitted to WHO for Phase II of the testing in Spring 2007 were slightly different from the nets submitted for Phase I. It was still a [Later Formula] recipe in terms of the three main additives but it contained an additional additive and the formulation of the polymer was different [533].

The application for WHOPES interim approval for the Later Formula nets

The progress of the Defendants’ application for WHOPES approval was as follows. Again, the numbers in square brackets refer to paragraph numbers in the Liability Judgment.

On 3 May 2005, Mrs Sig wrote to Dr Zaim of WHOPES requesting the new guidelines for evaluation. She referred to the fact that the company had just received the final report of the tests on nets at the Montpellier Lab and that they have started field trials in two countries. This must refer to the testing of Sample 19 which is a First Formula net [448]. In response, Dr Zaim sent Mrs Sig a pre-publication copy of the guidelines for testing LLINs. Later in May, Mrs Sig sent Dr Zaim a copy of the Montpellier First Formula Net Report together with some information about Netprotect [450].

In August 2005, Mrs Sig wrote to WHO asking them to evaluate the new LLIN Netprotect. She told them that they had taken a random selection of 20 nets ‘from a running production of 50,000 nets’. She also said that they had sent 15 nets to the Montpellier Lab and kept five nets for WHOPES reference [477]. At the same time, Mrs Sig wrote to the Montpellier Lab forwarding them 15 nets. Arnold J found that these nets had been made in the July 2005 production run and were therefore First Formula nets [481].

In September 2005, Dr Zaim wrote to Dr Skovmand and Mrs Sig saying that WHO had received a letter from VF informing it of legal action against Danish Intection [484]. Mrs Sig wrote to WHO enclosing a comparison sheet showing the differences between Netprotect and VF’s product PermaNet. She subsequently wrote to WHO telling them that approval would now be sought for the nets by IIC rather than by Danish Intection and that Danish Intection was going into liquidation [490].

In October 2005 Dr Skovmand wrote to WHO on behalf of IIC asking WHOPES to start evaluating Netprotect for WHOPES I and WHOPES II approval. He referred in this letter to the Montpellier Lab tests that had been run on the product and said that those tests demonstrated that the product was regenerating at a faster time than the most successful existing LLIN on the market. Dr Skovmand asked where he should send the nets and how many they needed. Arnold J commented that this showed that Dr Skovmand was aware that the tests carried out by the Montpellier Lab were not in accordance with the WHOPES I protocol but Dr Skovmand was hopeful that these results would persuade WHO to fast track the evaluation of Netprotect [491].

In November 2005 WHO wrote to Mrs Sig saying that they would evaluate the Netprotect product at the request of IIC, provided that assurances could be given that there was no dispute about the product with VF. Dr Skovmand wrote to WHO giving those assurances though, as the Judge noted, these assurances were not entirely accurate [493].

In December 2005 Dr Zaim wrote to Dr Skovmand indicating that the application was not complete because further specification and risk assessment was needed. On 24 January 2006 Dr Skovmand sent WHOPES the formal application for approval of the Netprotect product, enclosing a package of data including a specification, a toxicology evaluation and reports on the physical properties of the net samples (weight, tear strength and bursting strength). The specification enclosed described the net and said:

“Preliminary studies at LIN, Montpellier, has shown that the impregnation resist more than 20 washes without a decline in activity. These studies will be repeated under the new protocol for regeneration nets”

Further, under the heading ‘Release or retention index of active ingredient’ the specification said:

“A trial in 2005 with Net Protect showed that after exposure to three washes in a day, the net regenerated within two days as measured in mortality data where the mortality stopped around 90 % and did not develop further in the next 12 days. The net was stored at 30°C between bioassays. Appendix II

The first test with Net Protect carried out by this laboratory showed that the net could be washed 30 times within 2 – 3 days interval before exhaustion when stored at 40°C between washes (Appendix III). [REDACTED]”

It is accepted by the Defendants that that description comes from the Montpellier First Formula Nets Report. The specification does not give the recipe for the nets so there was nothing to alert WHOPES to the fact that the results being cited might relate to nets made to a different recipe from the ones to be sent for testing.

In January 2006 WHO invoiced IIC for US$13,000 for WHOPES I testing. This shows that by this date enough information had been provided by IIC to satisfy WHO that they could accept the Netprotect product for evaluation. Further material was sent by Dr Skovmand to Dr Zaim including toxicity data from Tagros, the suppliers of the deltamethrin.

On 26 March 2006 Dr Skovmand sent Dr Zaim an email saying that he had received nets from production to send to WHO and asking where to send them. Dr Zaim replied that they should be sent to the Montpellier Lab. The Judge accepted that samples had indeed then been sent to the Montpellier Lab in April 2006 and that these samples had been to the Later Formula [513].

On 13 September 2006 WHO wrote to Dr Skovmand enclosing the results of the WHOPES I final report by the Montpellier Lab and a risk assessment by the Finnish Institute of Occupational Health. The WHOPES I Report records that the nets had a regeneration time of 3 days and met the criteria to progress to WHOPES II testing.

In Spring 2007 more samples were submitted for Phase II testing. As I have mentioned, these samples were still made using the Later Formula but were slightly different in other respects from the sample 114 net that had been sent to the Montpellier Lab in April 2006 [533].

Between 10 and 13 December 2007 WHOPES held a meeting at which it considered the results of the Phase I and Phase II tests and decided to grant WHOPES interim approval. This decision was notified to the Defendants in March 2008.

Conclusion on quasi-consultancy fee

The history of the development of Netprotect and of the process by which the product attained WHOPES interim approval show, in my judgment, that substantial use was made by the Defendants of the First Formula which Arnold J comprised misuse of VF’s confidential information. This is not a situation where the initial research taken from the Fence database was rapidly superseded by Dr Skovmand’s subsequent research. It is true that the Defendants had to arrange and pay for a great deal of testing of First Formula nets, Later Formula nets and nets made to other formulae over many months. There is also no doubt that Dr Skovmand had to bring his considerable expertise to bear in analysing the results of the tests and gradually determining which formulae would give improved performance. I bear in mind in particular:

On the other hand, as I discussed earlier in relation to the royalty for some of the sales of First Formula nets, there was a substantial amount of work done by Dr Skovmand in pursuing possible changes to the recipe which turned out not to be fruitful and there were other aspects of the recipe, including other additives, changes to the polymer composition and to the extrusion temperatures which contributed to the success of the final product.

I also bear in mind when considering an appropriate quasi-consultancy fee that VF paid Dr Skovmand approximately $1.8 million between 1999 and 2005, of which nearly $493,000 was paid in 2003 [329]. These figures show that companies in this sector expect to pay substantial fees for access to expertise for developing these products.

Taking all these factors into account I consider that a quasi-consultancy fee of US$150,000 is an appropriate reflection of the use made of the First Formula nets and of tests results on those nets, when balanced against the further work carried out and the expertise brought to bear when arriving at the approved Netprotect product.

Damages for accelerated entry

There are two stages in determining whether any damages are payable for accelerated entry. The first is to work out how much time was saved in the development of Netprotect by Dr Skovmand’s use of VF’s confidential information at the start of the development process. The second is to work out how much of that time translated into accelerated sales of Later Formula Netprotect nets on the market.

How much time was saved by the Defendants’ use of VF’s confidential information?

The parties disagreed on how long it would have taken for Dr Skovmand to come up with the Later Formula without using VF’s confidential information. This is in part because they do not agree on how much of the information used by Dr Skovmand was confidential. VF assert that the identity of the additives used in the Fence database and an understanding of the effect of various additives on the speed of migration of the insecticide to the surface of the net after washing were all secret and could not legitimately have been used by Dr Skovmand to develop a net after leaving VF.

My task at this stage of the proceedings is not to make fresh findings as to what material was or was not confidential but to award damages for the infringement that Arnold J found had taken place as set out in the Liability Judgment. At paragraph [648], Arnold J recorded that it was common ground that the scope of any obligation of confidence owed by Dr Skovmand to VF did not extend beyond the information that amounted to trade secrets of VF’s once Dr Skovmand had ceased to work for VF. Dr Skovmand was entitled after the termination of his consultancy to use information forming part of his own skill, knowledge and experience even if it was learnt during the course of that consultancy. The Judge then considered the well-known case of Faccenda Chicken Ltd v Fowler [1987] CH 177 on the distinction between those two categories of information. Applying the various factors which the case law indicates are relevant, Arnold J held ([654]) that the information in the Fence database primarily consisted of two sorts of information: (i) recipes for PE yarns and nets and (ii) the results of tests on such yarns and nets. The specific information which VF contended had been misused consisted of the recipes and results for particular samples. VF also relied upon the information which could be derived from the results recorded in the database, such as what they reveal about the effect of [Additive A] and [Additive C] on [REDACTION]. The Judge noted that it has frequently been held that secret formulas and secret manufacturing processes are precisely the kind of information which can in appropriate circumstances constitute trade secrets. His view was that experimental results and deductions drawn from experimental results, particularly concerning products under development, fall into the same category. He went on to hold that the information in the Fence database was separable from Dr Skovmand’s general skill, knowledge and experience. This was particularly true of the recipes. He therefore held that that the information contained in the Fence database, and in particular the recipes contained in that database, constituted VF’s trade secrets.

The Judge also held that what he called the ‘Ciba recommendations’ were confidential to VF and were not part of Dr Skovmand’s knowledge that he was free to use as he pleased. This refers to information arising from discussions between Dr Skovmand and a representative of Ciba in 2001 and 2002 when Dr Skovmand was working for VF. The two men had discussed the likely properties of different Ciba additives and the Ciba representative had made recommendations about the concentrations that should be tried. These recommendations were later refined in the light of feedback from Dr Skovmand, about what additives to use: see paragraphs [174]–[182] and [643].

Arnold J’s conclusion was that the information contained in the Fence database and in particular the recipes and the bioassay results for particular samples constituted VF’s trade secrets. Those secrets had been misused by Dr Skovmand in that he used them to identify which recipes were worth repeating and to decide what variations should be tested.

In the light of this, I do not accept VF’s submission that the Judge found that the trade secrets covered the very identity of the three main additives such that in the counterfactual world Dr Skovmand must be assumed not to know that it would be useful to incorporate [Additives A, B and C] in the polymer. Such a finding would have been contrary to the evidence the Judge set out concerning the patents filed by VF in January 2002 in relation to its insecticidal tarpaulin called Zerofly: [187]. The Judge noted that that specification disclosed the use of [REDACTED]. The specification also states that [REDACTED]. I note here that VF did not assert that the use of deltamethrin for a bed net was confidential.

In my judgment the right question is how long it would have taken Dr Skovmand to arrive at the First Formula assuming that he knew that it was worth while experimenting with the three main additives but he would not have known the proportions to use. He would have known that it was possible that the additives not only performed the function for which they were being sold generally by Ciba but that they might also have an effect on the [REDACTED].

The Defendants say that it would have taken Dr Skovmand a matter of weeks to come up with the First Formula. This, they argue, is a case where Dr Skovmand merely cut a corner at the outset of his work but could very easily have got there without doing so. He would have drawn up an experimental grid of different proportions of the different additives and bioassayed those. I do not accept that this would have involved only a few weeks’ work. Certainly it took much longer than that when Dr Skovmand was working for VF. The subsequent experience of the Defendants when they were developing Netprotect shows that the process of producing samples and testing them is rarely smooth-running as there may be problems with the quality of production of the yarn samples at the factory and with the efficacy of the bioassays which may need to be repeated. The assumption that in 7 weeks a full 21 washes could be achieved is not borne out by what happened in the real world. I consider that Mr Larsen’s and Dr Skovmand’s evidence about the costs and time involved in coming up with the First Formula is over-optimistic.

For VF Dr Pates-Jamet provided a timeline for developing a LLIN and came up with a figure of between 21 to 30 months, taking into account that some of the steps can be carried out at the same time. However, her evidence is not so helpful because the question here is not how long it takes to produce a rival net with WHOPES interim approval, but how much of a short cut was made by the Defendants in using VF’s confidential information at the start of the process for developing Netprotect. Further, her first step taking 6 – 12 months includes time taken for the identification of the proper additives whereas I do not consider that Dr Skovmand would need to start from scratch in trying to identify the three main additives as the best candidates to test.

VF also argue that the counterfactual world should assume that Mr Larsen would not be involved in the project until the expiry of his restrictive covenant. Because his involvement was key to the project, in particular in encouraging investors to fund the work of Danish Intection and Bestnet, the start date for the counterfactual should be September 2005 rather than April 2004. I do not accept that is right. Certainly in constructing a counterfactual world one generally assumes that the parties will behave lawfully. However, VF have already brought proceedings against Mr Larsen for breach of his restrictive covenant and have been awarded damages for that breach by the Danish court. If the breach of the restrictive covenant is now to be treated as accelerating Netprotect’s entry to the market by 15 months, that may result in substantial additional damages being payable by Mr Larsen effectively for the same breach of contract. In the particular circumstances of this case, therefore I do not consider it is right to assume that the process of devising the First Formula in the counterfactual world would only start in September 2005.

VF further rely on evidence that the Defendants’ financial position was precarious at the start of the development period. Mrs Sig’s evidence and the evidence of Dr Skovmand on urgency that I have already quoted, indicate that there was serious concern about the continued viability of the company. However, I do not accept that there was a real risk that the delay caused by Dr Skovmand having to devise the First Formula without relying on the Fence database results would have caused the whole project to collapse. It is clear that in the real world the investors were prepared to tolerate substantial delays and set backs in the performance of the bioassays and other tests, in the WHOPES approval process and in the start of sales. As I have described in the Annex about the numbers of First Formula nets sold, Mr Johnsen’s company Syd Finans was prepared to step in to buy up the Karur Stock Nets to ensure that the liquidation of Danish Intection was not insolvent at this early stage of development.

Doing the best I can in the light of the evidence of the parties and the history of the development of the First Formula and the subsequent experience of Netprotect when conducting tests on net samples, I have concluded that the misuse of VF’s confidential information as found by Arnold J saved the Defendants about six months of work in the early stages of developing Netprotect.

Did that time saving result in additional sales of Later Formula nets?

Given that I am considering here damages for sales of Later Formula nets, the next issue is whether, if it had taken the Defendants an additional six months to arrive at the Later Formula and sample 114, that would have delayed the attainment of WHOPES interim approval. I have already set out how the Defendants’ application for WHOPES approval progressed. From that time line one can see that there were various pauses in the process begun in August 2005 when Mrs Sig first wrote to WHOPES up to the approval of the Netprotect product at the WHOPES meeting in December 2007. Should one just assume that approval would have been given in June 2008, six months later, or would some of the six month period have been be absorbed by pauses in the approval process anyway?

The first break in the WHOPES time line was the false start in August 2005 when Mrs Sig attempted to start the evaluation process using samples of First Formula net. This attempt was abandoned because of VF’s intervention informing WHO of the pending litigation between VF and the Defendants. It was only later that year, in November or December 2005, that Dr Skovmand’s company IIC reopened discussions with WHOPES about the application for approval and it was in January 2006 that WHOPES accepted the application. The Defendants argue that three or four months should be regarded as absorbing part of the additional six months taken to arrive at the Later Formula.

On this point I accept the submissions of VF that it should not. Dr Skovmand did not stop his development work when Mrs Sig sent the samples to WHOPES in August 2005. He continued working on improving and testing the product and, as we know, the product ultimately submitted by him in April 2006 was not the same as Mrs Sig had submitted. It is not possible to say whether a First Formula net would have obtained WHOPES interim approval as it was never subjected to those tests. The Later Formula samples were not ready before March 2006 and could not have been submitted before April 2006 even if there had not been the false start in August 2005. It is right therefore to assume in the counterfactual world that the submission of samples for WHOPES I testing would have taken place six months later in October 2006 rather than April 2006 if Dr Skovmand had had to devise the early stages of the development process himself without misusing VF’s confidential information.

I therefore turn to the question of when would interim approval have been granted if the WHOPES application and samples had been lodged in October 2006 instead of April 2006.

I have set out earlier in this judgment a passage of evidence from Dr Skovmand describing the urgency surrounding the submission of the application for WHOPES approval in March 2006: see paragraph 132 above. He explained there that the reason he sent off samples to a recipe that had not yet been tested was his concern that several other manufacturers might gain a place in the queue before Bestnet and cause a substantial delay to Bestnet’s approval. However, it appears that this fear was not well-founded and I do not consider that a delay of six months in lodging the application would have resulted in a more substantial delay to the start of WHOPES I testing. Mr Grantham and Mr Østergaard describe the entry of competitors into the market over the years. It appears that only BASF entered the market before 2009 so that the delay would not have had the consequences that Dr Skovmand feared at the time.

Evidence about the timings of the different phases of WHOPES testing both in general and in the specific case of Netprotect was given by Dr Pates Jamet for VF. The difficulty is that Dr Pates Jamet’s evidence as to what generally happens do not tally with what happened in the real world. Thus, it is common ground that the WHOPES I testing at the Montpellier Lab started in April 2006 shortly after the samples of Later Formula net were submitted by the Defendants. Dr Pates-Jamet’s evidence is that Phase I testing will take at least 60 days for a net with a regeneration time of 3 days (that is the maximum regeneration time permitted by WHO). One would have expected therefore the WHOPES I results to be reported to the Defendants in around June 2006. However it appears that it was only on 13 September 2006 that WHO wrote to Dr Skovmand with the Montpellier Lab’s report on the Phase I evaluation and testing. That report is not dated so it is not clear when the testing was carried out and the report completed and hence why it appears to have taken much longer than Dr Pates Jamet suggests is the normal period.

The letter from WHO of 13 September 2006 also enclosed the Finnish Institute of Occupational Health’s toxicology report. This states that it was submitted to WHOPES on 8 September 2006. It may be therefore that WHOPES waited until it had received the favourable toxicology report before informing the Defendants of the results of Phase I and that it only received that report shortly before the letter of 13 September 2006. In the light of that, I consider that it is fair to assume that the time taken for the production of the Phase I WHOPES test report and the toxicology report would have been the same in the counterfactual world as it would have been in the real world. I assume therefore that WHOPES I testing and toxicology reports would have been completed six months later in March 2007 rather than September 2006 as happened in the real world.

As to the timing of the WHOPES II testing, Dr Pates Jamet’s evidence is that this takes around three to six months. This covers “washing the nets, getting everything ready, running the study (which takes about 6 weeks) as well as time to analyse the results and write up the findings”. In the real world we know that Dr Skovmand submitted further samples (to a slightly different recipe) for WHOPES II testing in Spring 2007. What is not clear is why the Defendants or WHO waited from September 2006 until Spring 2007 before starting the Phase II testing if that is indeed what happened. It is therefore difficult to determine whether that pause in the process should be added on to the time taken in the counterfactual world. In other words, should I assume that if WHOPES I had been completed in March 2007 rather than September 2006, there would then have been a six month gap until September 2007 before WHOPES II field tests commenced or should I assume that WHOPES II testing would have started shortly after WHOPES I testing was completed? The evidence on this point comes from Dr Pates Jamet who states:

In the light of that I conclude that if the WHOPES I test results had only been available in March or April 2007, it is still likely that the WHOPES II testing would have occurred at the same time it occurred in the real world, during the mosquito seasons in Burkina Faso and Tanzania. There would have been no need for a pause of five or six months between the two phases of testing to await the 2007 mosquito season. I recognise that the period of delay in the real world between 13 September 2006 and March 2007 was used by the Defendants further to refine the recipe for Netprotect and that the samples submitted for WHOPES II testing in March 2007 were to a different formula from those subject to WHOPES I testing. But there is no reason to suppose that the WHOPES I version of the Later Formula would have performed less well in the field tests than the WHOPES II version of the Later Formula nets.

Further, even if the Phase II testing had been slightly delayed, there was still a period of delay in the real world between the conclusion of those tests and the WHO evaluation meeting that took place between 10 and 13 December 2007.

I find therefore that even if it had taken Dr Skovmand an additional six months to arrive at the Later Formula, avoiding using VF’s confidential information, it would not have made any difference to the date on which WHOPES interim approval was achieved. There was therefore no accelerated entry onto the market for the Later Formula nets. No damages in respect of Later Formula nets are payable under this head.

In the light of the findings set out above my conclusions in summary are as follows:

This Annex sets out my findings on the disputes about various sales of nets and to what formula those nets were made.

The Karur stock of 61,800 nets at November 2005

In his 19th witness statement, Mr Larsen referred to a document originating in India attached to an email dated 21 November 2005 from the Indian manufacturer of nets, Siva to Mrs Sig. This set out the stock that Siva held as at that date. The stock (‘the Karur Stock Nets’) was described as:

This stock count was supported by two ‘certificates’ issued by Siva, one dated 15 September 2005 and one dated 13 October 2005 showing slightly more than that amount of nets being put into stock in Karur. Mr Larsen’s evidence was that the nets recorded in the stock count on 21 November 2005 are the same nets as recorded in the 15 September and 13 October certificates and that the small discrepancy is the result of pilferage or some nets being used as samples. Mr Larsen said that ‘the nets in question would all have been made to the unsatisfactorily low density of about 25 gm/m2’. This meant that you were able to tear them apart with your hands. This was not because Siva had made mistakes in the production but that the recipe to which Intection wanted to make nets for sale had advanced to ensure a better quality net. He accepted that these nets would have been made to the First Formula.

The parties do not agree about what happened to the Karur stock nets. Mr Larsen’s evidence was that they were sold to three companies called Inmoya, Alpinter and Amadis (‘IAA’) because those companies are purchasers of returns and low quality products. The 61,800 nets were therefore not additional sales of First Formula nets for which VF is entitled to be compensated, because the sales of the nets to IAA are already included in Mrs Sig’s spreadsheet, albeit that he now accepts that those sales were of First Formula nets and not of Later Formula nets, as shown in the spreadsheet.

VF dispute that the sales of nets to IAA were the Karur stock nets. They say that these are additional nets that must have been sold to someone but that the documents evidencing those sales have not been disclosed and are not included in Mrs Sig’s spreadsheet.

Mr Larsen was frank that his conclusion that the Karur stock nets were sold off cheaply is not the result of his recollection of the events at the time but rather the result of his inferences from:

In cross-examination, Mr Platts-Mills put to Mr Larsen various countervailing documents which indicate that the 61,800 Karur Stock Nets are not the same nets as were sold to IAA.

First he showed, and Mr Larsen agreed, that a substantial number of the Karur Stock Nets, that is 40,419 nets, were in fact sold to a company called Syd Finans which is controlled by Nikolai Johnsen, one of the investors in the Defendants. Syd Finans is not a net-selling company so would not have marketed the nets themselves. But if the 40,419 nets were sold to Syd Finans, how could they also be sold to IAA? Mr Larsen attempted to explain this by a rather tortuous account. He said that when it became clear that Danish Intection (to whose order Siva had made the nets) was going to be put into liquidation, he was concerned that this should be a solvent liquidation so that the company’s creditors should not be harmed. Therefore the following transaction took place:

The difficulty with marrying this account of events with Mr Larsen’s earlier evidence is that it means that Bestnet must have sold the nets to IAA at the time that the nets were supposedly owned by Syd Finans. Mr Larsen then produced a document which he said had been provided to him by Syd Finans, drawn up from information in its corporate documents, setting out in a table the acquisition of the nets and the sales of the nets back to Bestnet over a period of 18 months or so from 21 October 2005 until 15 November 2007. This document was provided by Mr Johnsen to Mr Larsen during the course of the preparation for the Inquiry Trial (‘the Syd Finans table’).

Mr Larsen said that when Bestnet started to make sales of cheap product, they sold off that part of that Karur Stock Nets that had not been transferred to Syd Finans (that is 61,800 less 40,419 = 21,381) because those still belonged to Siva and Siva needed the money more than Syd Finans did. The 61,800 Karur Stock Nets never actually left the Karur factory and were not physically separated by Siva into those nets owned by Syd Finans and those nets owned by Siva (having given credit for them to Intection).

This explained, he said, why the sizes of the nets supposedly included in the 61,800 Karur Stock Nets did not marry up with the sizes of the nets sold to IAA.

More importantly, the Siva email describing the Karur Stock Nets clearly says that all the nets in stock were 156 size mesh (that is they had 156 holes to the square inch) whereas the invoices to Inmoya clearly state that the nets being sold were mesh size 232 and nets sold to Amadis were mesh size 200. Mr Larsen’s evidence was that when Danish Intection first started production of Netprotect nets they were producing two mesh qualities, 156 holes per square inch and 232 holes per square inch. The higher mesh size indicates that the net is denser and heavier. Later they moved to mesh sizes 136 and 200 because they decided to use thicker yarn and thicker yarn cannot be knitted up so densely.

Mr Larsen’s explanation was simply that these mesh sizes on the invoices had ‘crept in’ to the invoices. He also had to accept that his assertion that the nets weighed only 25 g/m² might not be right once it was pointed out to him that in her submission to WHOPES in September 2005 Mrs Sig told WHOPES that the Netprotect nets weighed 31g/m². It also seems to me unlikely that even a buyer of ‘seconds’ quality nets would be prepared to buy nets which pull apart in your hands as Mr Larsen described.

I accept that Mr Larsen believes that he has done his best to come up with an explanation which he thinks fits the documents and his recollection, such as it is. However, when the documents contain information which is inconsistent with his account of events, he simply brushes that aside saying that the document is incorrect and cannot be relied on in that respect although he does rely on other aspects of the document in so far as it evidences facts which support his theory. When he was challenged with aspects of the email describing the Karur Stock Nets numbers and the Syd Finans table that were inconsistent with his account, he fell back on assertions about how long ago these events were; how casual record keeping had been at the time and so forth.

I find that he has reconstructed events starting with the assumption that the 61,800 Karur Stock Nets were sold to IAA and therefore do not constitute substantial unaccounted for sales of First Formula nets. Given that his evidence starts from the proposition which it sets out to prove (that is that there are no additional, unaccounted for sales of First Formula nets) I cannot rely on it as an explanation. There are too many inconsistencies between that account and the documents. In particular it seems that the inconsistent net sizes and mesh size are an insuperable obstacle for Mr Larsen’s account.

Similarly there was no reason to put any mesh size on the invoices to IAA if, as Mr Larsen claims, IAA were indifferent to the sizes and mesh of the nets. The idea that Bestnet would send an invoice to Inmoya for a specific mesh size and then supply nets which they had been told were a different mesh size, not knowing whether all or some of them were the wrong mesh size seems an unnecessarily risky way to do business. If they did not know what the stock sizes were one would expect Siva to leave those details blank in its Stock Note or put ‘various’ and similarly one would expect to see similar wording used in the invoices to IAA. It is much more plausible that the net sizes and the mesh details were accurate in the stock note and that the invoices sent to IAA were also accurate in describing the size and mesh size of the nets being sold. This means that the nets sold to IAA cannot have been from the Karur Stock Nets at all.

There are some additional points which lead me to reject Mr Larsen’s explanation:

On balance, my conclusion on the Karur Stock Nets is that they were First Formula nets which are not the same nets as were sold to IAA. They were different nets which must have been sold by Bestnet during the early months of Bestnet’s business but for which there is no documentation – or at least none that has been found and disclosed by the Defendants. I reject Mr Larsen’s explanation as being inconsistent with the contemporaneous documents.

The large IDA Order

Mrs Sig’s spreadsheet shows at line 15 that there was a sale to the IDA Foundation (‘IDA’) of 121,828 nets for a total price of US$487,312 under invoice dated 12 June 2006. There were two factual issues to be resolved about the sales of nets to IDA. The first is whether there were in fact two sales both of that number of nets between December 2005 and May 2006 or only one sale and the second is what recipe were the nets made to.

As regards the number of sales, Mrs Sig’s and Mr Larsen’s evidence was that there was only one sale during this period. There was an initial order for 110,000 nets given at a meeting with IDA in December 2005. This was later revised to 121,000 and finally to 121,828 nets. The number eventually supplied was 121,828 nets.

VF accept that the order progressed in that manner but they point out that the Bestnet invoice dated 12 June 2006 gives the product identification code for 121,828 light blue nets whereas the order from IDA dated 28 March 2006 for 121,828 nets gives a product code relating to white nets.

This discrepancy in the colour code is not enough to convince me that there were in fact two orders from IDA for 121,828 nets and I find that there was only one order.

Turning then to the question as to what recipe was used to make these 121,828 nets, that depends on when they were in fact manufactured. They are shown on Mrs Sig’s spreadsheet as being made to the Later Formula but as I have described, that was simply because the invoice date was later than the date on which the Later Formula nets started to be made at the Karur factory in March 2006. Mr Larsen accepted at the Inquiry hearing that the invoice date might not be an accurate reflection of when the nets were made and hence to which recipe they were made.

At the Inquiry Hearing Mr Larsen carried out a different exercise to try to work out when these nets were likely to have been made. He did this by calculating how much masterbatch was available for making nets before the instruction was given to the factory to start using the WHOPES I recipe rather than the First Formula. Any sales made after that instruction was given in about March 2006 and after new masterbatch supplies were ordered must have been made to the new recipe.

Mr Larsen used the following elements to make his calculation as to whether the IDA nets were from the First Formula nets or the Later Formula nets. He notes that as at January 2006 there were the 61,800 Karur Stock Nets already discussed. He also knows that there was 725 kg of [Additive B] masterbatch that had been ordered on 16 December 2005. The IDA sale of 121,828 nets is at line 15 of Mrs Sig’s spreadsheet. He therefore runs through invoices on lines 1 – 14 treating those which he believes were supplied from the Karur Stock Nets as depleting those stocks and treating those which must have been made new as depleting the 725 kg of [Additive B] masterbatch. He explains how he calculated how much [Additive B] masterbatch is used up by the manufacture of each net. This of course differs according to the size of the net and needs to take into account that there is some wastage of net (he assumes 8 per cent of the total net used for a particular order) and that the nets have seams that use up centimetres of net not reflected in the overall measurements of the net.

The sales shown on Mrs Sig’s spreadsheet as being earlier than the large IDA Order are as follows:

Thus when he arrives at the IDA sale at line 15, Mr Larsen has calculated that Siva has only [REDACTED] kg of [Additive B] masterbatch left at the Karur factory. This is not enough to make 121,828 nets because it takes [REDACTED kgs of [Additive B] masterbatch to make that number of nets. There is evidence that a new delivery of masterbatch had taken place in April 2006, in time to make the IDA 121,828 nets to the Later Formula. There is an email from Siva dated 20 April 2006 in which Siva asks Alok to supply 13,300 kg of masterbatch. He asks for 1,800 kg of that masterbatch to be delivered immediately. That 1,800 kg includes 275 kg of [Additive B]. This 275 kg plus the remaining [REDACTED] kg is still not enough, Mr Larsen acknowledges, to make up the 121,828 nets. He concludes that further deliveries out of the 13,300 kg order would have been needed to complete the IDA order. But there is no evidence of any such further delivery being made in time to enable delivery of the nets in May 2006. Mr Larsen therefore accepts that an alternative explanation is that some of the nets for the IDA order were made earlier, to the First Formula. At most, he calculates that about 24,000nets out of the 121,828 might have been made earlier to the First Formula. But according to his recollection of how the large IDA order was handled, he concludes that the most likely answer is that all the 121,828 nets were made using the left over [REDACTED] kg plus the 275 kg immediately despatched masterbatch from the April order and then however much more of the April 2006 order of 13,300 kg of masterbatch was needed to complete the order.

VF regard the exercise carried out by Mr Larsen as entirely bogus. In cross-examination the line of questioning started by challenging the idea that the sales to Inmoya, Alpinter and Amadis at lines 5, 6 and 13 of the spreadsheet or the later sales to those companies were sales from the Karur Stock Nets. I have already set out why I find that these were not in fact the sales of the Karur Stock Nets. If the 17,000 nets sold in lines 5, 6, and 13 were new nets made to the First Formula then the whole exercise of depleting the [Additive B] masterbatch by the sales must be very wrong because the factory would have used up an additional [REDACTED] kg of masterbatch to make those new First Formula nets before it got to the IDA sales in line 15. Mr Platts-Mills pointed out to Mr Larsen in cross-examination that the list provided by Alok as to the deliveries of masterbatch to Siva over the period 27 April 2006 to May 2007 shows that no [Additive B] masterbatch was delivered to Siva between April and October 2006. The October 2006 supply would have come far too late to complete the IDA order. So there must have been some deliveries of [Additive B] masterbatch between the delivery of 725 kg in December 2005 and the delivery in April 2006 in order to complete the IDA order. Mr Larsen accepted that this must be true.

VF had a counter suggestion as to where the extra masterbatch had come from to make the IDA nets prior to April 2006 and the IAA nets between December 2005 and April 2006, assuming that those IAA nets were not supplied from the Karur Stock Nets. They point to a pro forma invoice for masterbatch dated 30 December 2005. This, they say, shows that at that time, Siva was buying 23.750 tonnes of masterbatch – enough in fact to make [REDACTED] nets to the First Formula. They therefore suggest that there are substantial sales of nets, most likely First Formula nets, that were made from this masterbatch and which are not revealed in the invoices so far disclosed.

Mr Larsen said that this pro forma was not a real order for masterbatch but a document used to obtain a government licence which entitled Alok to refrain from charging certain local taxes on the supply of masterbatch because the masterbatch was to be used to make products which would ultimately be exported and hence be exempt from those local taxes. The practice was, Mr Larsen said, to produce a pro forma invoice for a very large amount of product. That enabled a licence to be set up for that value and then future supplies over the coming years would ‘use up’ the value of that licence as and when they were made. VF’s contention was that the document shows that a quantity of masterbatch was being bought which evidences not only that the IDA and IAA nets could easily have been made to the First Formula but that many thousands more nets may have been too.

On this point I accept Mr Larsen’s evidence that the pro forma invoice does not evidence the actual supply of such a huge quantity of masterbatch as at December 2005. The other documents show that masterbatch was ordered by Siva from Alok in small amounts sufficient to cover current manufacture and that Alok was able to deliver the masterbatch very swiftly indeed when the need arose. There could be no justification for suddenly ordering and taking advance supplies of such large volumes of masterbatch. There were a few large orders for nets but nothing over 100,000 except for the IDA order. It seems to me very unlikely that all this masterbatch was to satisfy much larger orders for nets so soon after Bestnet had launched the product and before it had WHOPES interim approval. However, although I accept that this pro forma invoice performed the function that Mr Larsen describes, there must have been some undocumented call off of these amounts between December 2005 and April 2006 to make the IDA and the IAA nets. Unfortunately the schedule of supply provided by Alok for the purposes of these proceedings only starts in April 2006.

Further, there is other evidence that the IDA nets were made well before April 2006. On 6 January 2006 Mrs Sig emailed Siva telling them that a total of 121,000 nets were to be produced for the IDA order. This was followed by a purchase order from Bestnet to Siva for 121,000 nets on 13 December 2005. Of course if the nets were made then, they would have been made to the First Formula. There is then an invoice from Siva to Bestnet for 121,828 nets dated 20 February 2006 to be paid for by letter of credit number DOCLC 24439 dated 4 January 2006 with Jyske Bank Denmark. There is then evidence that on 20 February 2006 Siva indeed wrote to its local bank in Karur presenting documents evidencing shipment for the full 121,828 nets and asking them to forward those documents to Jyske Bank.

Mr Larsen accepted the account up to that point. But he denied that all the nets for the IDA order were in fact made then. He says that after the oral order was given by IDA at a meeting in December 2005 – originally for the smaller number of 110,000 nets – Bestnet started to gear up to produce them. But then finalising the actual order dragged on because of delays in getting approval for the funding in the Central African Republic. The formal order did not arrive until very much later. He accepts that Siva was representing to the Bank that it had made and shipped all the nets in February 2006 in order to unlock the letter of credit. His evidence was that this was a subterfuge, agreed between Siva and Bestnet, to enable Siva to claim all the money under the letter of credit even though they had not in fact yet made or shipped the nets. He described this as Siva ‘telling his bank a little white lie’ because Siva was running out of funds; all Bestnet’s money was tied up in the letter of credit and so to help Siva out, Bestnet agreed to Siva claiming all the letter of credit funds prematurely.

Again it seems to me that Mr Larsen’s evidence is based on a possible reconstruction of events which makes an assumption about the very proposition he is trying to establish. The question is: when were the IDA nets made? He is convinced that they were not made until 30 May 2006. Therefore he has to come up with a reason why all the documents point to their having been made much earlier. He accepted that he did not actually remember when the IDA nets were made. The reason he is convinced that the nets were not made until the end of May is that the invoice sent by Bestnet to IDA for the nets shows that the goods were delivered on 30 May 2006 and the usual practice of Bestnet was to send out the invoice as soon as the nets had been manufactured. That, he says, is the most accurate date as to when the order was made.

Mr Larsen’s evidence on this is implausible. It seems unlikely that a company with limited cash resources would go to the expense of setting up a letter of credit as the mechanism for paying Siva for goods but then allow Siva to claim all the money before having made or shipped the goods. It also put the bank at risk by requiring them to pay out before the goods were made or shipped, as there does not seem to be any documentary evidence supporting the suggestion that Siva acted with Bestnet’s connivance on this. I accept that Mr Larsen remembers that the final order from IDA was much delayed but in my judgment it is much more likely that the December and February documents mean what they say. The nets were all made by then and there was then a wait for the order to come in. The invoice could not of course be sent out before the order came in. The actual order from IDA for the 121,828 nets is not among the documents so the precise date of that is not known.

Mr Larsen’s reliance on the invoice delivery date as showing the IDA nets were made in May 2006 is also inconsistent with the evidence he gave in his 16th witness statement where he clearly says that the nets for IDA had been made in March 2006.

I therefore find that the 121,828 nets for the large IDA order recorded at line 15 of Mrs Sig’s spreadsheet were made to the First Formula.

Sales in lines 4 to 14 of Mrs Sig’s spreadsheet

I have described already the other sales in the earlier lines of Mrs Sig’s spreadsheet and Mr Larsen’s evidence about them. As regards the early sales to IDA (lines 1, 2 and 3 amounting to 11,500 nets), the sales to ICRC (line 4 sale of 20,000 nets and line 7 sale of 5,000 nets) and the very small sales (lines 8 and 9 sales of 24 nets), Mr Larsen accepted that these should be accounted for as First Formula nets.

As regards sales to IAA (lines 5, 6 and 13 amounting to 17,000 nets) Mr Larsen’s evidence was that these were taken from the Karur Stock Nets so he accepted that they were First Formula nets for this purpose. As I have found that they were not from the Karur Stock Nets, I must then come to a conclusion about whether the IAA sales were First Formula or Later Formula nets. There are also still disputes about smaller sales to IDA (lines 10, 11 and 12 amounting to 10,000 nets) and the sale to Clause Bogh (line 14 sale of 1000 nets).

I bear in mind the very unsatisfactory way in which the exercise of showing which nets were made to which recipe was carried out by the Defendants in purported compliance with the court’s orders. Given my findings in relation to the large IDA order, I consider that it is likely that all the sales recorded on Mrs Sig’s spread sheet up to and including the line recording IDA’s order of 121,828 were of nets made to the First Formula. Thereafter they were nets made to the Later Formula or subsequent formulae.