Merck KGaA v Merck Sharp & Dohme LLC & Ors

This judgment was handed down remotely at 10 a.m. on 19 September 2025 by circulation to the parties or their representatives by email and by release to the National Archives Mrs Justice Bacon: Introduction

1. This is an inquiry into damages following a protracted dispute between the parties concerning the rights to use the word “Merck” in the UK. The parties, which are both multinational companies in the life sciences and healthcare sectors, have a common heritage and were historically both run by members of the Merck family. I will refer to the claimant as Merck and the defendants collectively as MSD .

2. The present litigation commenced in 2013, and has been the subject of judgments of the High Court and Court of Appeal as summarised further below. Ultimately, following a trial and further hearing before Norris J, MSD was found to have breached a 1970 agreement between the parties, and infringed Merck’s marks, by the use of the word “Merck” in the UK. On 28 July 2020 Norris J granted various declarations and injunctions, and ordered an inquiry as to damages.

3. Merck seeks damages on the basis of a reasonable royalty under a notional (fictional) licence negotiated between the parties, permitting MSD to perform acts otherwise prohibited by contract or by Merck’s trade mark rights. On that basis its pleaded claim was for damages of approximately £50.5m. By the start of this hearing that had reduced to £46.1m. That headline figure relied on Merck’s analysis of comparable licences. For the reasons discussed further below, by the closing submissions Merck’s focus was on an alternative economic benefits approach to the licence fee valuation, which it contended indicated a damages figure of around £18.7m.

4. MSD’s position was that the present case is not appropriate for an award of licence fee damages at all, whether calculated on a comparables approach or an economic benefits approach. In the alternative, MSD contended that Merck’s comparables analysis was fatally flawed, leaving the economic benefits approach as the only meaningful basis for assessment. Calculated on that basis, MSD’s submission was that the appropriate licence fee calculation should be no more than around £2.2m plus interest.

5. At the hearing, Mr Brandreth KC and Mr Ivison represented Merck; Mr Hobbs KC and Mr Hollingworth represented MSD. Shortly after the hearing the parties provided brief further submissions on the issue of the discount rate, in light of the parties’ positions in their closing submissions. On 5 September 2025 the parties provided, at the request of the court, some further calculations on specific figures forming part of the notional licence fee calculation. Witnesses Merck’s witnesses of fact

6. Merck relied on the evidence of four witnesses of fact. Ultimately MSD only sought to cross-examine one of them, Ms Vittinghoff, and did not challenge any part of the evidence of the other three witnesses.

7. Jonas Koelle has worked for Merck since 2002 and has been head of its trade marks department since 2014. He provided two witness statements for this trial addressing, in particular, the nature and extent of the conduct of MSD regarded as objectionable by Merck, his involvement with Merck’s rebranding exercise from 2013 onwards, the various trade mark licences granted by Merck (including within the Merck group of companies on transfer pricing terms), and how Merck would have viewed the prospect of granting a licence to MSD to cover MSD’s acts amounting to breaches of contract and trade mark infringements in the UK.

8. Katrin Menne joined Merck in 2014 as a project manager for its rebranding exercise, and is now head of brand and content marketing. She provided a short witness statement with evidence about Merck’s use of social media and internet marketing, and an estimate of the costs to Merck of maintaining two mirrored websites (one for the USA/Canada and another for the rest of the world).

9. Frank Sielaff is an external consultant to Merck, who previously worked at Merck between 2005–2019, eventually becoming director and head of its digital reputation and web platform department. He provided a short witness statement addressing Merck’s use of internet analytics tools (specifically Google Analytics, SEMrush and SimilarWeb) to gather and compare web traffic data for Merck’s websites as well as competitor sites, including MSD websites.

10. Undine Vittinghoff has been the director of Merck’s tax department since 2010. In 2014 Ms Vittinghoff co-authored a report for Merck analysing the company’s umbrella brand from a transfer pricing perspective. She provided a witness statement on the background and purpose of that report, and was cross-examined on this (by Mr Hobbs) for almost half a day. While her witness statement was given in English and her cross-examination was also conducted in English, she was occasionally assisted by a German interpreter. She was a straightforward witness who gave helpful answers that were not materially challenged by Mr Hobbs. MSD’s witnesses of fact

11. MSD relied on the evidence of five witnesses of fact, all of whom were cross-examined by Merck.

12. Anita Parfitt has worked for MSD since 2016 and is now its international data strategy and innovation lead, specialising in data analytics across MSD’s websites, social media and other promotion channels. She provided two short witness statements addressing the regulatory framework governing the promotion of pharmaceutical products in the UK, MSD’s UK marketing activities including its use of websites and social media, and the way in which the list price of products is determined. She was cross-examined briefly (by Mr Ivison), during which she gave apparently straightforward answers, albeit that she frequently responded that the questions put to her concerned matters falling outside her role, which she was therefore unable to answer. Ultimately her evidence was not very useful for the issues for determination in this case.

13. Kyra Lanza has worked for MSD since 2010 in various roles, and is now an Associate Vice President dealing with MSD’s ESG strategy and engagement, responsible for MSD’s company brand at an enterprise level, including consideration of how that brand is perceived by the general public. She provided a short witness statement addressing the changes made to MSD’s websites, email addresses and social media sites as a result of these proceedings, as well as commenting on MSD’s marketing strategy during that period. Ms Lanza was cross-examined for about an hour (by Mr Brandreth) on MSD’s geo-blocking and the company’s migration of email addresses in the UK. It became apparent that her witness statement had provided information about the costs of Merck’s email migration, in circumstances where that information was not within her knowledge but had been supplied by her solicitors Linklaters, and was materially inaccurate. She was highly defensive when asked about this, and displayed considerable reluctance to engage with the problems with this part of her evidence. Unfortunately, it was clear that her witness statement comments on this issue were not her own evidence. I therefore do not regard her as an entirely reliable witness, and treat her evidence with some caution.

14. Michael DeFerrari has worked for MSD since 2000, particularly on the “Merck Manuals”, a medical reference work. He provided two witness statements for this trial, giving an overview of the costs associated with the production and maintenance of the manuals. He was briefly cross-examined (by Mr Ivison) on the reputational benefits to MSD of the manuals. He was an honest and straightforward witness, and there was no material challenge to his evidence.

15. Peter Young has been the director of commercial operations at the fourth defendant since 2019. He provided a witness statement which covered similar ground to that of Ms Parfitt in relation to MSD’s animal health business, explaining the regulation of sales of veterinary pharmaceuticals in the UK, the company’s UK marketing activities including its use of websites and social media, and changes made to the company’s UK websites, email addresses and printed material as a result of these proceedings. He was cross-examined briefly (by Mr Brandreth). His answers were somewhat defensive and did not always engage with the questions.

16. Martin Rogers has worked for MSD since 2012 and has been its financial controller since 2015. He provided two very short witness statements addressing some specific points relating to MSD’s financial documents and promotional expenditure on the names MSD and Merck Sharp & Dohme in the UK. He was briefly cross-examined (by Mr Brandreth), in particular on the figure given by him for MSD’s promotional expenditure. His answers acknowledged that his evidence on this point was not complete and did not explain the basis for the expenditure figure provided by him. Expert evidence

17. Both parties relied on expert evidence as to the quantum of damages based on a notional licence granted by Merck to MSD. Merck’s expert was Andrew Wynn , a senior managing director at FTI Consulting, where Mr Wynn works as an economist and financial analyst. He has over 25 years of experience in providing economic and financial advice, including significant experience of conducting financial analysis in intellectual property disputes. MSD’s expert was Jeffery Stec , a managing director at Berkeley Research Group, where he leads its intellectual property practice. He also has over 25 years of experience in economic consulting, including advising on trade mark disputes and computation of damages in various forms.

18. Each expert submitted an initial expert report followed by a second (reply) report. Following a meeting between the experts, a joint statement was produced setting out the areas of agreement and disagreement. Each expert then produced a 10-page “expanded summary” of their position on the areas of disagreement. These were useful documents for the purposes of pre-reading, but ultimately the cross-examination of the experts focused on the detail of their analysis which was set out in the far longer individual expert reports.

19. Mr Wynn favoured the valuation of a notional licence by reference to comparable licences. That approach was robustly criticised by Dr Stec in his evidence, as being entirely unreliable. Dr Stec’s view was that the only reliable basis for assessment of damages was an analysis of the economic benefits and costs to the parties, comparing the position as it was with the position had the breaches of contract and trade mark infringements not occurred. Mr Wynn’s view was that the economic benefits approach did not fully capture all the relevant factors in the hypothetical negotiation, so he carried out that analysis in the alternative only by way of a cross-check.

20. As described in detail further below, Mr Wynn eventually accepted in his cross-examination by Mr Hollingworth that his comparables analysis was based on a study that was unreliable and did not produce a statistically meaningful result. It is very unfortunate that this acknowledgement came so late in the day, because it completely undermined Mr Wynn’s evidence on this point. In the light of that belated concession, I do not consider that I can place any weight on Mr Wynn’s comparables analysis.

21. That left the economic benefits approach favoured by Dr Stec. Mr Wynn had considered that approach in the alternative, albeit that this was not the approach that he preferred. His calculation of economic benefits was greater than that of Dr Stec, but it was still far below his assessment based on his comparables analysis. Merck therefore sought to uplift the figure produced by the economic benefits approach, by reference to evidence and analysis not provided by Mr Wynn, but which Mr Brandreth sought to introduce through his cross-examination of Dr Stec.

22. I was, however, not willing to permit the cross-examination of Dr Stec on the basis of evidence and analysis not discussed between the experts, nor considered previously by Dr Stec, nor explored with Mr Wynn in his oral evidence. Where there has been a lengthy expert report process, including a meeting of experts and a joint expert statement, the court is entitled to expect that the parties’ cases, on the issues considered by the experts, will be presented on the basis of the evidence that emerges through the expert report process. As noted in my judgment in Cabo v MGA [2025] EWHC 1451 (Ch) , §§49–50, the joint meeting of the experts provides an opportunity for the experts to revise their opinions as appropriate, and the court expects the experts to engage properly and objectively with the evidence of the other side. If an expert fails to do so, but then changes their position on a particular point during the course of cross-examination, the court must then determine the case having regard to that change of position. The party relying on that expert cannot expect the court to allow it to attempt to salvage its position by advancing an entirely new case at that stage, on the basis of an analysis that has not previously been considered by either of the experts in their reports.

23. I have, therefore, approached the economic benefits assessment on the basis of the evidence of Dr Stec and Mr Wynn, as set out in their expert reports and explored in their cross-examination. On that issue, I consider that both Dr Stec and Mr Wynn provided evidence that was generally helpful, albeit that they were at times both somewhat defensive in their cross-examination. Factual and procedural history

24. The background to the present proceedings has been set out extensively in the previous judgments, in particular the trial judgment of Norris J [2016] EWHC 49 (Pat) and the subsequent Court of Appeal judgment [2017] EWCA Civ 1834 . What follows is a short summary. The Merck companies

25. The claimant in these proceedings, Merck, is the successor to a family business under the name of E. Merck, which started as an apothecary’s shop in Darmstadt, Germany in 1668. The business was very successful and grew over the years, establishing its own manufacturing facilities and expanding to various other countries. In the late 1800s a member of the Merck family established another business in the USA, which was initially economically supported by the German Merck business. The American business was incorporated as Merck & Co, Inc, and the present defendants are the successors to that business. The American business also became very successful and expanded its operations outside the USA.

26. Following the confiscation of property that took place as a result of the First World War, the American business became an independent company, but the two businesses cooperated over the use of the word “Merck”. Broadly speaking, that word was used by the American business in the USA and Canada, and by the original German business in the rest of the world. In 1932 the companies entered into a formal agreement along those lines, but it was challenged by the US Department of Justice as being contrary to the Sherman Anti-Trust Act, and was cancelled in 1945 by a consent decree in the New Jersey District Court.

27. In 1953 Merck & Co, Inc merged with Sharp & Dohme. By then both the American and German companies were multinational businesses, and both were seeking to develop their businesses globally using the Merck trade mark. The 1955 and 1970 Agreements

28. On 12 September 1955 the companies signed a new agreement (the 1955 Agreement ) under which, in broad terms: i) E. Merck recognised the exclusive right of Merck & Co to the use of the Merck trade mark in the USA and Canada. In turn Merck & Co agreed that E. Merck could refer to itself as “E. Merck A.G.” or “Emanuel Merck offene Handelsgesellschaft” in the USA and Canada, provided that those names were geographically identified with Germany. ii) Merck & Co recognised the exclusive right of E. Merck to the use of the Merck trade mark in Germany. In turn E. Merck agreed that Merck & Co could refer to itself as “Merck & Co” or “Merck Sharp & Dohme”, again provided that those names were geographically identified with the USA or Canada (in the case of Merck & Co) or a country other than Germany (in the case of Merck Sharp & Dohme).

29. In all other countries, it was agreed that E. Merck was entitled to use the word Merck as a trade mark, and would not object to the use by Merck & Co of Merck Sharp & Dohme as a trade mark or name; nor would it object to the use by Merck & Co of the name of Merck & Co where combined with a geographical designation identifying it with the USA or Canada.

30. In addition, in relation to countries other than the USA, Canada and Germany, clause 7 of the 1955 Agreement provided: “In all other countries Merck & Co shall promptly and in any event no later than three years after the effective date of this agreement cancel all existing registrations, withdraw all applications and discontinue all use of the trademarks Merck, Merck Cross and MerckMerckMerck.”

31. In 1970 a new agreement was signed (the 1970 Agreement ) which modified the 1955 Agreement to reflect a change in Merck’s corporate name. The main terms of the agreement remained the same as under the 1955 Agreement. Reflecting clause 7 of the 1955 Agreement, clause 7 of the 1970 Agreement provided: “In all other countries Merck & Co has undertaken to cancel all existing registrations, withdraw all applications and discontinue all use of the trademarks ‘Merck’, ‘Merck Cross’ and ‘MerckMerckMerck’.”

32. The 1970 Agreement, together with a protocol agreed in 1975, still governs the relationship between the parties, and formed the basis of Merck’s contractual claim in these proceedings. The present proceedings and previous judgments

33. In 2009 MSD merged with Schering-Plough. Merck thereafter became concerned at the expansion of the use by MSD of the word Merck outside the USA and Canada, including in various websites with global reach, such as merck.com, merckformothers.com and merckmanuals.com.

34. Eventually on 8 March 2013 Merck filed its claim in these proceedings, alleging breach of clause 7 of the 1970 Agreement and infringement of its UK and international registered trade marks for the word MERCK.

35. In 2014 there was a hearing before Nugee J of a preliminary issue as to whether the governing law of the 1970 Agreement was German law or New Jersey law. In his judgment of 21 November 2014 [2014] EWHC 3867 (Ch) (the preliminary issue judgment ), Nugee J found that the governing law was German law.

36. The trial of the action then took place before Norris J in April 2015, and judgment was handed down on 15 January 2016: [2016] EWHC 49 (Pat) (the trial judgment ). Norris J found that MSD had breached the 1970 Agreement and had infringed Merck’s registered trade marks, in ways that were more than de minimis . While he found that those registrations should be revoked in respect of some of the goods and services in the specifications, on the grounds of non-use, that did not affect his findings of infringement. The final order following the trial was settled on 3 March 2016 following a hearing in February 2016.

37. Both Merck and MSD appealed. MSD appealed against the judge’s findings on both breach of contract and trade mark infringement, as well as part of his conclusion on revocation of the Merck trade marks for one of the specifications, and the terms of the final order. Merck, for its part, appealed on more limited grounds, concerning some aspects of the relief ordered by the judge, and the judge’s finding that the use by MSD of merck.com as a domain name and @merck.com email addresses was not a trade mark infringement.

38. The appeal was heard in June 2017, and judgment was given by Kitchen LJ, with whom Patten and Floyd LJJ agreed [2017] EWCA Civ 1384 (the Court of Appeal judgment ). The court upheld the findings of breach of contract and rejected the argument that these were no more than de minimis . However, it upheld Merck’s appeal against the finding of trade mark infringement, on the grounds that the judge had addressed the issue of infringing uses of the word Merck in terms that were too general and insufficiently reasoned, and that some of his findings were internally inconsistent. The court also upheld MSD’s appeal on the scope of the revocation of Merck’s trade marks, and found that the formulation of the final relief did not allow MSD the opportunity to be heard and was not explained by further reasons.

39. In light of the conclusions of the Court of Appeal, four matters were remitted to the High Court for rehearing: (i) the issue of partial revocation of the registered trade marks; (ii) whether the impugned activities of MSD constituted use in the UK in the course of trade in relation to any relevant goods or services; (iii) if so, whether those uses were more than de minimis ; and (iv) the appropriate form of relief to be granted for both the breach of contract claim and the trade mark infringement claim.

40. The hearing of the remitted matters took place before Norris J in July 2018, and judgment was given by Sir Alastair Norris on 20 May 2020, following the judge’s retirement and a period of illness: [2020] EWHC 1273 (Ch) (the remitted matters judgment ). The trade mark infringement claim was, this time round, considered by reference to 32 samples of the use of the word Merck by MSD. The judge concluded that 23 of those samples had infringed Merck’s trade marks in a way that was more than de minimis . On the revocation point, he revoked certain of the specifications for Merck’s marks, replacing them with narrower specifications. His final order (dated 28 July 2020) granted an injunction restraining further breaches of the 1970 Agreement and infringements of Merck’s trade marks, and ordered an inquiry as to the damages (if any) suffered by Merck as a result of MSD’s breaches of contract and trade mark infringements. Alongside the order, Sir Alastair Norris gave a further judgment providing his reasoning for some of the specific terms of the order: [2020] EWHC 2120 (Ch) .

41. MSD breached the order of Sir Alastair Norris, leading to a judgment of Edwin Johnson J on 12 April 2024 finding a series of breaches to have occurred by the use of the word Merck by MSD in various documents, webpages, domain names and email addresses: [2024] EWHC 820 (Ch) . Other breaches alleged by Merck were, however, found not to have been established. MSD’s appeal to the Court of Appeal was dismissed: [2025] EWCA Civ 343 . Relevant findings of breach and infringement

42. For the purposes of the issues for determination in this trial, it is not necessary to set out a comprehensive summary of the findings of Norris J in his trial judgment (to the extent upheld by the Court of Appeal) and the remitted matters judgment, as to the acts which were infringements of Merck’s trade mark rights and/or breaches of contract. Rather, it is sufficient to give a flavour of those findings.

43. First, it was not alleged that MSD had ever used the word Merck on the packaging of any goods sold or offered for sale in the UK. But Norris J did find there to be “direct use” of the Merck mark in the sense that some of Merck’s webpages, accessible in the UK, used the word Merck in a way which he found was use in a trade mark sense to designate Merck as the originator of MSD’s pharmaceutical products, and was therefore a trade mark infringement (as well as a breach of contract).

44. These included (for example) a www.merck.com webpage promoting oncology drugs, bearing the logo “MERCK Be well”, referencing further information regarding “Merck oncology clinical trials” and “Merck’s investigational medicines access policy”, and explaining that the trade marks on the website were “owned, licensed to, promoted or distributed by Merck”. Norris J found that the references to Merck reinforced the link between the products and the sign MERCK: remitted matters judgment §§41–42.

45. Similarly, in relation to MSD’s services, another page on the www.merck.com website promoted the role of “Merck Clinical Trials” in developing new drugs, identifying enrolling locations in the UK for clinical trials. That page likewise included a Merck logo and made repeated references to “Merck”. Norris J held that the Merck sign was being used in relation to goods and services in the course of trade, because the “whole object of the page is to promote confidence in the products currently associated with the Merck US sign … these pages are part of a suite of sites, one function of which is to associate in the mind of the reasonable consumer (i) goods sold and services provided by MSD in the UK with (ii) the ‘MERCK’ sign”: remitted matters judgment §43.

46. Norris J also considered various examples of webpages referring to the online version of the Merck Manuals, the originator of which was, he noted, a business using the sign Merck. Although the Merck Manuals are a not for profit service, he found that they were used in the course of commercial activity promoting products that were sold by MSD in the UK (and referred to elsewhere in the Merck suite of websites), such that the sign Merck was used in relation to goods or services in the UK: remitted matters judgment §§55 and 57.

47. Further examples included pages from the www.merck-animal-health.com website with references to Merck Animal Health as well as the Merck Manual for Pet Health. Norris J found that the object and effect of the pages was to associate the products and services mentioned with the Merck sign and thereby to promote the business of the MSD Animal Health business unit: remitted matters judgment §68.

48. In addition to the use of the Merck sign on websites, Norris J found that users of the www.msd-uk.com website were frequently directed to @merck.com email addresses. That was a trade mark infringement as well as a breach of contract: trial judgment §§116, 124, 134; remitted matters judgment §§84–86. Norris J also considered that specific emails sent in response to an enquiry from a doctor about a particular MSD product were infringing use, on the basis that the emails referred to the product as a Merck product. That was found to be direct dealing with a consumer in the marketplace, using the Merck mark: remitted matters judgment §82.

49. It was common ground that all of the conduct found to amount to a trade mark infringement was also a breach of contract. Some of the conduct objected to by Merck was, however, only found to be a breach of contract and not a trade mark infringement. An example of that was a set of YouTube pages linking to videos about various medical conditions, with the videos described as being “by Merck”, with the Merck US logo displayed. Norris J considered that these were “at one remove from product promotion”: remitted matters judgment §76. Merck had also objected to a small number of offline uses of the Merck sign, including presentations at conferences and press releases. These were breaches of contract but were not found to be trade mark infringements: remitted matters judgment §§78–81.

50. As to MSD’s overarching approach, Norris J commented that: i) The examples before the court were “the exemplification of a policy to bring before the market in the UK the sign ‘Merck’ in relation to the products of Merck US at every opportunity (with the consequence that the impact of Merck Global’s exclusive use of the ‘Merck’ mark in the UK was diluted)”: remitted matters judgment §89. ii) “Since 2009 Merck US has demonstrated a determination to push the boundaries as far as it considers desirable in the interest of its business, and to set those boundaries entirely according to its perceptions. … Merck US continued to regard itself as entitled to refer to itself as ‘Merck’ in the UK. … Merck US did not act dishonestly. That does not mean that the impugned acts were oversights or accidents. The impugned acts were part of a conscious policy”: remitted matters judgment §99. iii) “… the evidence also shows a determination to continue to link ‘MSD’ with the ‘Merck’ identity whenever possible”: remitted matters judgment §100. Issues

51. It was not disputed that MSD’s breaches of contract and trade mark infringements had caused at least some loss to Merck. The disputes were as to how that loss should be quantified in the circumstances of this case, the reliability of the evidence as to the different possible approaches to quantification, and the extent to which Merck had properly pleaded the main approach on which it ultimately relied. On those points the parties’ arguments raised the following specific issues: i) Whether it is appropriate to award damages in the present case based on a notional licence fee. ii) If so, whether there is any reliable basis for assessing the value of that notional licence by reference to the royalty rate contained in Merck’s intragroup licences (i.e. the comparables analysis). iii) If the comparables analysis is in principle reliable, what adjustments should be made to the Merck intragroup royalty rate to reflect the specific circumstances of the notional licence in the present case, and across what period should the notional royalty be applied? iv) Whether Merck can, on its pleaded case, rely on an economic benefits approach to value the notional licence, as an alternative to the comparables approach. v) If so, how the notional licence should be quantified based on the evidence of economic benefits. Relevant law The principle of licence fee damages

52. The availability of licence fee damages, sometimes described as “negotiating damages” or “user damages”, for certain types of infringement of intellectual property rights and breaches of contract, is well-established in the case-law. An early example in relation to a patent infringement was the case of Watson, Laidlaw & Co v Pott, Cassels and Williamson (1914) SC (HL) 18; 31 RPC 104, where the question was whether the claimants could recover damages for sales of infringing products in a territory where the claimants could not themselves have traded. Lord Shaw considered that they could, under a principle which he described as “price or hire”. In a passage at p. 31, he noted that this principle applied not only to patent cases, but also “wherever an abstraction or invasion of property has occurred”. He endorsed the approach adopted by Fletcher Moulton LJ in Meters v Metropolitan Gas Meters (1911) 28 RPC 157, where it was said at p. 165: “The reward to a patentee for his invention is that he shall have the exclusive right to use the invention, and if you want to use it your duty is to obtain his permission. I am inclined to think that it would be right for the court to consider what would have been the price which – although no price was actually quoted – could have reasonably been charged for that permission, and estimate the damage in that way.”

53. In Morris-Garner v One Step [2018] UKSC 20 , Lord Reed (with whom Baroness Hale, Lord Wilson and Lord Carnwath agreed) described the principle as applying not only to patent infringements but also to breaches of other intellectual property rights of a proprietary character (§§4 and 26). Following a long discussion of the previous case-law, he concluded that the principle could also apply, in certain circumstances, to breaches of contract. Those circumstances were (as described at §§91–92) ones in which: “… the loss for which compensation is due is the economic value of the right which has been breached, considered as an asset. The imaginary negotiation is merely a tool for arriving at that value. The real question is as to the circumstances in which that value constitutes the measure of the claimant’s loss. … such circumstances can exist in cases where the breach of contract results in the loss of a valuable asset created or protected by the right which was infringed, as for example in cases concerned with the breach of a restrictive covenant over land, an intellectual property agreement or a confidentiality agreement. … The claimant has in substance been deprived of a valuable asset, and his loss can therefore be measured by determining the economic value of the asset in question. The defendant has taken something for nothing, for which the claimant was entitled to require payment.”

54. Lord Reed’s conclusions on the availability of negotiating damages, set out at §95, included the following points in particular: “(1) Damages assessed by reference to the value of the use wrongfully made of property (sometimes termed ‘user damages’) are readily awarded at common law for the invasion of rights to tangible moveable or immoveable property … The rationale of such awards is that the person who makes wrongful use of property, where its use is commercially valuable, prevents the owner from exercising a valuable right to control its use, and should therefore compensate him for the loss of the value of the exercise of that right. He takes something for nothing, for which the owner was entitled to require payment. (2) Damages are also available on a similar basis for patent infringement and breaches of other intellectual property rights. (3) Damages can be awarded under Lord Cairn’s Act in substitution for specific performance or an injunction, where the court had jurisdiction to entertain an application for such relief at the time when the proceedings were commenced. Such damages are a monetary substitute for what is lost by the withholding of such relief. (4) One possible method of quantifying damages under this head is on the basis of the economic value of the right which the court has declined to enforce, and which it has consequently rendered worthless. Such a valuation can be arrived at by reference to the amount which the claimant might reasonably have demanded as a quid pro quo for the relaxation of the obligation in question. The rationale is that, since the withholding of specific relief has the same practical effect as requiring the claimant to permit the infringement of his rights, his loss can be measured by reference to the economic value of such permission. (5) That is not, however, the only approach to assessing damages under Lord Cairn’s Act. It is for the court to judge what method of quantification, in the circumstances of the case before it, will give a fair equivalent for what is lost by the refusal of the injunction. … (10) Negotiating damages can be awarded for breach of contract where the loss suffered by the claimant is appropriately measured by reference to the economic value of the right which has been breached, considered as an asset. That may be the position where the breach of contract results in the loss of a valuable asset created or protected by the right which was infringed. The rationale is that the claimant has in substance been deprived of a valuable asset, and his loss can therefore be measured by determining the economic value of the right in question, considered as an asset. The defendant has taken something for nothing, for which the claimant was entitled to require payment.”

55. Earlier in the judgment, in a passage relied on in particular by MSD, Lord Reed acknowledged the difficulties that inevitably arise in the assessment of the fee resulting from a hypothetical negotiation, and the limitations of that approach: “74. It is also necessary to recognise that the assessment of a hypothetical release is itself a difficult and uncertain exercise. In cases such as Wrotham Park, Bracewell v Appleby and Jaggard v Sawyer , judges estimated in a rough and ready way the amount which the claimant might fairly and reasonably have demanded as a quid pro quo for the relaxation of the obligation in question. More recently, the practice has developed of instructing forensic accountants to give expert evidence about a hypothetical negotiation between a reasonable person in the position of the claimant and a reasonable person in the position of the defendant. Such imaginary negotiations have become increasingly elaborate, and a host of questions can emerge as to the basis on which they should be hypothesised. …

75. The artificiality of the exercise can be a further problem. Since the aim is to arrive at an objective valuation, the fact that the claimant might in reality have been unwilling to release the defendant from the obligation is not necessarily a problem … But the premise of the hypothetical negotiation – that a reasonable person in the claimant’s position would have been willing to release the defendant from the obligation in return for a fee – breaks down in a situation where any reasonable person in the claimant’s position would have been unwilling to grant a release, as was found to be the position in Marathon Asset Management LLP v Seddon …”

56. At §75, Lord Reed was making the point that it is not a bar to the application of licence fee damages that, on the evidence, one or both parties might not actually have entered into the agreement that is being postulated. A similar point was made by Lord Sumption in his judgment in Morris-Garner , at §107, noting that “the assumption of a willing buyer and a willing seller, acting reasonably, means that one is not trying to reconstruct what the particular parties would hypothetically have done”. Lord Sumption cited in this regard the comments of Lord Walker in Pell Frischmann Engineering v Bow Valley [2009] UKPC 45 , §49, that: “It is a negotiation between a willing buyer (the contract-breaker) and a willing seller (the party claiming damages) in which the subject-matter of the negotiation is the release of the relevant contractual obligation. Both parties are to be assumed to act reasonably. The fact that one or both parties would in practice have refused to make a deal is therefore to be ignored.”

57. The postulation of a hypothetical negotiation, as a tool for the assessment of the value of the asset lost or compromised, does not therefore require a finding that the parties in question would or might actually have carried out such a negotiation, but rather asks whether such a negotiation could reasonably have taken place. Lord Reed’s comments at §75 recognised that there might be cases where such a negotiation cannot reasonably be posited, such as the situation in Marathon Asset Management v Seddon [2017] EWHC 300 (Comm) , [2017] FSR 36 , to which Lord Reed referred. In that case, Leggatt J observed: “234. There are other cases, however, in which it is unrealistic to value the benefit obtained by the defendant in this way because the defendant could not reasonably have expected to purchase a licence from the claimant for its activity … A patent, trademark, copyright or confidential information may be a means of generating revenue for the owner of the right through the granting of licences … But in other cases it may be relied on to protect intellectual property which the owner of the right would not be willing to allow anyone else to exploit at all – or not at a price which would rationally be paid for a licence.

235. In cases where the defendant could not reasonably have expected to purchase a licence from the claimant for its use of the claimant’s property (nor to obtain an equivalent benefit in another way) it in my view makes no sense to value the benefit by postulating a hypothetical negotiation between a willing seller and a willing buyer. Such a method makes no sense because in such a context the negotiation is not merely fictional in the sense that it did not actually happen but fictional in the stronger sense that it lacks any verisimilitude. The law ought not to employ fictions of the latter sort.”

58. That is no doubt why Lord Reed’s conclusions emphasised that a valuation by reference to a hypothetical negotiation is one possible method for quantifying damages for the use wrongfully made of property, but is not the only possible approach; and that it is for the court to judge in each case what method of quantification will give a fair equivalent for the value of the loss. The relevant counterfactual

59. There was some discussion in the parties’ submissions as to the correct approach to the counterfactual in a case where the court is applying a licence fee approach. The starting point is that in a case where the damages payable in respect of an intellectual property infringement or breach of contract are quantified on a licence fee basis, the court postulates a hypothetical situation where the relevant infringing acts were instead carried out pursuant to a licence or agreed release of obligation. That was explained by Lord Shaw in Watson, Laidlaw & Co at p. 32: “…the cardinal question which remains always to be answered in these infringement suits [is] the question put by Vice-Chancellor Page Wood in Penn v Jack, viz: ‘What would have been the condition of the plaintiff if the defendants had acted properly instead of acting improperly. That condition if it can be ascertained, will, I apprehend, be the proper measure of the plaintiff’s loss.’ To apply the principle: The appellants did this Java trade improperly. Had they done it properly they would have done it under royalty. That royalty the respondents would have obtained.”

60. In assessing the value of the notional licence or contractual release that would have been negotiated, the court will of course have regard to the existing rights and obligations of the parties. That is perhaps an obvious point: the court cannot ascertain the value of a release of a particular contractual obligation without considering the extent of the rights initially granted under the agreement said to have been breached. This is a specific example of the general principle that the hypothetical negotiation is assumed to have taken place on the basis of the commercial context existing at the time. As Arnold J noted in Force India Formula One Team v 1 Malaysia Racing Team [2012] EWHC 616 (Ch) , [2012] RPC 29 , §386, (ii): “The primary basis for the assessment is 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”.

61. Consistent with these comments, Newey J in 32Red v WHG [2013] EWHC 815 (Ch) , §32, noted that the court should have regard to the circumstances in which the individual parties were placed at the time of the hypothetical negotiation, since the hypothetical negotiation is designed to establish the value of the wrongful use to the actual parties, rather than as between hypothetical parties. Likewise, both Arnold J in Force India (at §436) and Newey J in 32Red (§§46–50), emphasised that the subject matter of the hypothetical negotiation should be the specific infringing conduct.

62. I set out these points because of the submissions of both parties (and Mr Hobbs in particular) on this issue. I do not, however, understand any of these points to be controversial; nor, ultimately, did anything appear to turn on these points when applied to the specific elements of the damages quantification in the present case. General approach to uncertainties in the evidence

63. Whatever the appropriate basis for the assessment of damages in this case, it was common ground that the court cannot in the circumstances of this case expect to be able to quantify the damages with precision, but should instead approach the assessment of damages with “the exercise of a sound imagination and the practice of the broad axe” (as per the comments of Lord Shaw in Watson, Laidlaw & Co , pp. 29–30). I summarised the general principles in that regard in my judgment in Anan Kasei v Neo [2022] EWHC 708 (Ch) , [2022] FSR 21 , §85, as follows: “(i) The assessment of what would have happened in the counterfactual case is not a matter susceptible of precise estimation: “one cannot expect much in the way of accuracy when the court is asked to re-write history”: Gerber v Lectra [1995] RPC 383 , 395. (ii) The court’s task is therefore to do the best it can with the material available to it: General Tire and Rubber Company v Firestone Tyre and Rubber Company [1975] WLR 819, 826; Original Beauty Technology v G4K Fashion [2021] EWHC 3439 (Ch) , §75. (iii) Where the defendant’s wrongdoing has created uncertainties, those should where necessary be resolved by making assumptions generous to the claimant: Gary Fearns v Anglo-Dutch Paint [2010] EWHC 1708 (Ch) , §70. Green LJ commented to similar effect in NTN Corp v Stellantis [2022] EWCA Civ 16 , §26 that where a claimant has a justiciable right the procedural and evidential rules governing the enforcement of that right should not be so onerous that it makes the right too hard to vindicate. (iv) The court should also have regard to the extent to which it was in the power of one or other party to produce evidence on a particular point: see the principle stated by Lord Mansfield in Blatch v Archer (1775) 1 Cowp 63, 65, cited with approval by Lord Bingham in Fairchild v Glenhaven [2002] UKHL 22 , [2003] 1 AC 32 , §13.”

64. The application of those principles should not, however, lead the court to adopt a valuation which is pure speculation, without any evidential support: see Arnold LJ on appeal in Anan Kasei v Neo [2023] EWCA Civ 11 , [2023] FSR 14 , §124. Appropriateness of licence fee damages in the present case

65. Merck’s pleaded claim is for licence fee damages, on the basis of a notional licence willingly negotiated by the parties, permitting MSD to perform acts otherwise prohibited by contract or by Merck’s trade mark rights. No other basis for the assessment of damages is pleaded, and Merck therefore properly accepted that its damages claim stood or fell with its claim for licence fee damages.

66. MSD contended that the circumstances of the present case are such that there is no juridical basis for the assessment of damages on the basis of a notional licence fee. That proposition was advanced in various different ways.

67. First, MSD contended that the availability of licence fee damages, as a means of quantifying an infringement of an intellectual property right, differs as between patents (where licence fee damages are a typical basis for a damages quantification) and trade marks. The high point of MSD’s submission in this regard was the following passage in the judgment of Jacob LJ in Reed Executive v Reed Business Information [2004] EWCA Civ 159 , [2004] RPC 40 , §165: “I am by no means convinced that the ‘user’ principle automatically applies in trade mark or passing-off cases, especially where the ‘mark’ concerned is not the sort of mark available for hire. The ordinary case is one that just protects goodwill. For damages to be awarded on the user principle is close to saying there is no damage so some will be invented. It is not the same sort of thing as having to pay for use of an invention (the basis of the user principle in patents).”

68. That statement must, however, now be read in light of the judgment of Lord Reed in Morris-Garner v One Step , making clear that the licence fee approach to quantification of damages will in principle be an appropriate approach where the loss can be defined as the economic value of the right which has been breached, considered as an asset. As Lord Reed repeatedly noted, this will typically be the case for breaches of intellectual property rights, as well as breaches of agreements concerning intellectual property rights.

69. There is no reason, in principle, why that should not apply to trade mark rights. Indeed, in Easygroup v Easy Live (Services) [2023] EWCA Civ 1508 , [2024] FSR 15 , in relation to a passing off action, Arnold LJ stated at §49 that “[g]iven that the juridical basis of the action is invasion of property, it naturally follows that damages assessed according to the user principle should be available”. That must apply equally to the infringement of a trade mark right.

70. Mr Brandreth acknowledged that there might be cases where the very nature of the right infringed was such that a licence would be entirely unrealistic, such that they fell into the category of cases described by Leggatt J in Marathon where any negotiation would be “fictional in the stronger sense that it lacks any verisimilitude”, or as Lord Reed put it in Morris-Garner , “where any reasonable person in the claimant’s position would have been unwilling to grant a release”. But that is (as Mr Brandreth pointed out) not the position for Merck’s trade mark rights in the present case, which are quintessential commercial property rights, for which a licence presents no inherent difficulty.

71. The second strand of MSD’s argument was that by licensing the Merck mark to MSD, Merck would have created a situation where the use of the mark misled the public, such as to render the trade mark liable to result in revocation of the registration of the licensed trade mark under s. 46(1) (d) of the Trade Marks Act 1994 . On that basis, MSD contended that Merck could not realistically have granted a licence to MSD.

72. That argument is misconceived. Section 46(1) (d) covers the situation where, in consequence of the use made of the trade mark by the proprietor or with the proprietor’s consent in relation to the goods or services for which it is registered, it is liable to mislead the public, particularly as to the nature, quality or geographic origin of those goods or services. In so far as that provision covers deception as to trade origin at all, it cannot be interpreted so as to preclude, entirely, all licensing of the trade mark at all, not least given that the provision itself recognises the possibility of use of the mark with the consent of the proprietor.

73. Indeed, as Mr Brandreth pointed out, Merck’s own evidence from Mr Koelle was that Merck did grant licences to use the Merck mark “in a wide variety of situations”, which included not only licensing within the Merck group, but also licences to unaffiliated companies in different contexts including collaboration with third parties on product development, agreements on sponsorship, contract manufacturing and advertising, and transitional arrangements on divestments of businesses by Merck. Mr Hobbs notably did not explain, in his submissions on this point, how his submissions on s. 46(1) (d) were to be reconciled with that evidence.

74. Nor did Mr Hobbs attempt to reconcile his submissions on s. 46(1) (d) with the fact that the effect of the 1955 and 1970 Agreements was to permit MSD to use the names Merck & Co and Merck Sharp & Dohme in territories where the parties had agreed that Merck had the exclusive rights to the use of the Merck trade mark. As Nugee J commented at §90 of the preliminary issue judgment, the agreement by E. Merck to permit Merck & Co to do certain things which it might otherwise have been able to stop was not conceptually any different from a licence.

75. The third strand of MSD’s argument referred to Merck’s evidence that it would not in practice have licensed the Merck brand to MSD, to cover the uses which amounted to trade mark infringements and breaches of the 1970 Agreement. Mr Koelle’s evidence was that Merck would not have done so because of the risk to its reputation and identity. Ms Vittinghoff likewise said in her oral evidence that the Merck umbrella brand would not “generally” be granted to an unaffiliated third party. On the basis of that evidence, Mr Hobbs argued that this was a case where no reasonable person in Merck’s position would have been willing to license MSD to use the Merck mark.

76. I do not accept that argument. Merck’s evidence was, undoubtedly, that it would not in practice have granted a licence to MSD to permit the conduct impugned in the present case. As set out above, however, it is not a bar to the assessment of damages on a licence fee basis that one or even both of the parties in question would not actually have granted the sort of licence that is postulated in the hypothetical negotiation. Rather, the question is whether the hypothetical negotiation could reasonably have taken place between the parties.

77. In the present case, given the evidence that Merck could and did license the Merck mark to a wide variety of affiliated and unaffiliated companies, in different situations, and the fact that the 1955 and 1970 Agreements specifically permitted MSD to use the Merck mark in territories where the parties had agreed that Merck had exclusive rights to that mark, it is clear that the parties could in principle have negotiated a licence of the Merck mark to cover MSD’s acts of infringement and breach. The evidence before the court was that Merck would not itself have done so; but there was no evidence suggesting that the negotiation of such a licence would have been unrealistic for any reasonable person in Merck’s position.

78. Finally, MSD sought to rely on the Agreement on Trade-Related Aspects of Intellectual Property Rights ( TRIPS ), which provides in Article 21 that the compulsory licensing of trade marks is not permitted. It was very difficult to understand this point, given that there is no question of a compulsory licence in the present case. Rather, all that is said by Merck is that a notional licence can be postulated as a conceptual tool to provide a framework for the quantification of the loss of a valuable asset created by its trade mark rights and the 1970 Agreement.

79. I do not, therefore, accept MSD’s submission that the award of licence fee damages is in principle inappropriate in the circumstances of the present case. The relevant question is rather whether there is before the court a reliable basis for the assessment of quantum on that basis. The assessment of licence fee damages: overview

80. Mr Wynn (for Merck) and Dr Stec (for MSD) were instructed to provide estimates of the royalties that would have been agreed between the parties in a hypothetical negotiation. The parameters of that assessment were agreed. In particular, the assumption was to be that Merck was a willing licensor and MSD a willing licensee; and the hypothetical negotiation should be assumed to have resulted in a licence that would cover the exact scope of MSD’s misuses of the Merck mark in the ways that were found to have constituted breaches of contract and trade mark infringements. The experts agreed that this would have been a licence of the Merck mark as an “umbrella brand”.

81. The experts agreed that the hypothetical negotiation should be assessed based on a reasonable person’s expectations as to the future, rather than with the benefit of hindsight; but they also agreed that hindsight information could (at least in some respects) be used as a proxy for those expectations.

82. The experts also agreed that, in principle, two approaches could appropriately be used to value the notional licence in issue in these proceedings: i) A comparables approach would seek to value the royalty which would have been agreed under the notional licence by reference to similar licences which were based on the same or similar benefits. ii) An economic benefits approach would seek to value the notional licence by reference to the incremental economic benefits expected to be obtained through the licensee’s use of the rights granted by the licence, and the costs to the licensor as a result of granting the rights under the licence.

83. Mr Wynn’s preferred approach in the present case was the comparables approach. On that basis he calculated damages in a range of between £18.6–46.1m. If he used the economic benefits approach as a cross-check, he put the damages figure at between £8.1–8.7m. Dr Stec fundamentally disagreed with the comparables approach and said that it was not possible to apply that methodology, even in a modified fashion, to reach a reliable damages figure. His preferred approach was the economic benefits approach, which he used to reach damages figures of £1.27m (based on an assumed 2010 negotiation date) or £1.44m (2007 negotiation date).

84. The opening submissions of the parties reflected the positions of their respective experts on this point. MSD maintained essentially that position in its closing submissions. With adjustments for inflation and the discount rate, but excluding interest, MSD’s final damages figures were £2m (2010 negotiation date) or £2.2m (2007 negotiation date).

85. As noted above, however, Merck’s position in closing submissions changed significantly as a result of the change of position of Mr Wynn during his cross-examination. Although Merck’s written closing submissions still retained nominal reliance on the comparables approach, Mr Brandreth’s oral closing submissions made no further comment on that approach, and instead solely addressed a reformulated case on the economic benefits approach, which put the damages figure at a total of more than £14.15m.

86. Since Merck did not explicitly abandon its reliance on the comparables approach, it remains necessary to address that first, before turning to the alternative economic benefits analysis. Comparables approach Mr Wynn’s analysis

87. Mr Wynn’s analysis of comparable licences placed primary reliance on the royalty rates set by Merck for transfer pricing purposes for its intragroup licensing scheme, under which Merck licenses the use of its umbrella brand to its subsidiaries. Under that scheme, Merck sets a royalty rate of 0.33% for the use of its brand by subsidiaries in its healthcare division.

88. That royalty rate was set on the basis of a benchmarking study by EY (the EY benchmarking study ), which was prepared in September 2019 for the purpose of determining an arm’s length remuneration for Merck’s intragroup licences, in accordance with the OECD Transfer Pricing Guidelines, following a global rebranding project for Merck’s umbrella brand which had been implemented from 2013 onwards. The EY benchmarking study was in turn the basis for a report by EY entitled “Transfer Pricing Documentation for the introduction of an Umbrella Brand license”, which set out the royalty rates adopted in Merck’s intragroup licences for its umbrella brand (the EY TP report ).

89. The EY benchmarking study and the EY TP report described the following methodology:

90. EY sought to assess an arm’s length price using the comparable uncontrolled price method, which referenced the amount charged in comparable uncontrolled transactions under similar circumstances. EY was, however, unable to find comparator agreements which dealt only with the licensing of the company name or logo. Instead, it identified a set of four licensing agreements which licensed both the licensor’s name/logo and product trade marks (set 1); and a set of six licensing agreements which licensed only the licensor’s product trade marks (set 2). i) EY calculated the lower quartile, median, and upper quartile percentage royalty figures for each of sets 1 and 2, and then took the difference between those figures for each set as denoting the percentage interquartile licence rates for the licensing of only the company name or logo, without any accompanying licence of product trade marks. That calculation produced the following figures: Difference between set 1 and set 2 Lower quartile -0.17% adjusted to 0% (see (iii) below) Median 0.66% Upper quartile 1.56% ii) Since the difference between the lower quartile rates produced a negative result (i.e. the lower quartile rate for set 2 was in fact higher than the lower quartile rate for set 1), whereas Merck’s umbrella brand was assumed to provide a financial benefit to the licensees within the Merck group, EY adjusted the lower quartile difference figure to 0%. iii) EY used those calculations to determine the licence rates to be applied for intragroup licences for Merck’s life science and performance materials business, and Merck’s healthcare business. In respect of the former, a licence rate of 1.11% was selected, as being the mid-point between the median and upper quartile difference figures. In respect of the latter, noting that umbrella brands are less important in the pharmaceutical industry, where product brands play a much stronger role, a licence rate of 0.33% was selected. As can be seen from the table above, that rate represented the mid-point between the assumed lower quartile difference figure of 0% and the median difference figure of 0.66%.

91. Mr Wynn’s first report recognised the limitations of the EY analysis, noting that its conclusions were drawn from calculating differences in the royalty rates between two sets of transactions, in circumstances where the licences in the two sets were not identical. His conclusion was, however, that the EY analysis adopted a “reasonable approach” given the limited information available on licences for umbrella brands in isolation.

92. Mr Wynn looked at various other types of licences as potential comparators, but concluded that they were not sufficiently comparable to use for present purposes. He therefore considered that the transfer pricing rate of 0.33% applied for Merck’s intragroup healthcare businesses was the most appropriate starting point for his analysis. His conclusion rested on observations that (i) the licences were formulated on the basis of an arm’s length principle; (ii) they had real-world implications in respect of the tax paid; (iii) Merck had an incentive to keep its transfer pricing rate as low as possible given that Germany is a comparatively high tax jurisdiction; and (iv) the rate established was based on the EY benchmarking study.

93. Mr Wynn’s reply report, after considering Dr Stec’s criticisms set out below, maintained that the 0.33% figure used in Merck’s intragroup healthcare licences was “strong evidence of an appropriate starting point for the rate that would have been considered for the use of the word [Merck]”, and rejected Dr Stec’s criticisms of the comparables approach. Mr Wynn’s expanded summary following the joint expert statement likewise maintained that Merck’s intragroup royalty rate was a reasonable starting point and could be used to inform a “reasoned and supportable assessment”.

94. Mr Wynn did not, however, simply take the 0.33% rate, but used it as a starting point and applied various adjustments to it. These included a series of discounts to take account of the specific circumstances of the case, including the scale of MSD’s misuses of the Merck mark during different periods, the fact that MSD did not need a licence to use the word Merck in the ways permitted by the 1970 Agreement, and the fact that MSD did not use all of the rights licensed in Merck’s intragroup healthcare licences – most notably it did not use the Merck mark alone on product packaging. Going the other way, Mr Wynn added a potential premium adjustment on the basis of the evidence (discussed above) that Merck would have been unwilling to license the word Merck to a competitor.

95. Mr Wynn acknowledged that his adjustments were necessarily subjective and uncertain. He therefore expressed the results of his adjustments as ranges applied to different periods, rather than precise percentages. His results also distinguished between the breach of contract claim and the trade mark infringement claim. That led to royalty rate ranges of between 0.13% and 0.47%, depending on the period and the type of claim.

96. As regards the royalty base, Mr Wynn calculated two sets of figures, one set including non-UK revenues, and a second set excluding non-UK revenues. His final calculations were a damages range of £24.3–46.1m if non-UK revenues were included, and £18.6–35.4m if non-UK revenues were excluded. That was the basis for Merck’s claim of £46.1m at the start of the hearing. The criticisms of Mr Wynn’s analysis

97. Dr Stec’s consistent position was that Mr Wynn’s comparables analysis was fundamentally flawed. He advanced numerous criticisms. Fundamentally, his position was that given the unique parameters of the hypothetical negotiation in the present case, there was no reliable comparable licence, and that no number of adjustments to a non-comparable licence would lead to a reliable royalty result.

98. Dr Stec noted that Mr Wynn’s starting point of 0.33% was taken from Merck’s intragroup healthcare division licences, which were in turn based on the results of the EY benchmarking study. His criticism of that was that the sample sizes in the two sets of licences used by EY to calculate “arm’s length” royalty rates (as described at §88 above) were so small that the comparator rates calculated on the basis of the differences between the two licence sets were statistically meaningless. An economically and statistically valid analysis using the approach adopted by EY would, he said, require: i) a sufficiently large number of representative data points to generate a statistically robust analysis; as well as ii) the identification of two groups of licences which differed systematically only in the characteristic of interest (i.e. the licensing of the umbrella brand) but whose characteristics were otherwise very similar.

99. On the first of those two points, Dr Stec noted that the EY benchmarking study did not provide an analysis of whether the difference between the two sample sets was statistically significant. He therefore carried out his own analysis using the standard T-test methodology for measuring statistical significance. His conclusion was that the differences between the means of the two groups was not statistically significant. As he put it: if the T-test fails to show any statistically meaningful difference between the means of the two groups, then it cannot be concluded that the set 1 licence agreements included any premium on the royalty rate relative to the set 2 agreements. He reached the same conclusion using an alternative statistical test.

100. Dr Stec’s conclusion was therefore that the difference between the two sets could not, in statistical terms, be distinguished from zero. In other words, the comparator licences identified by the EY study provided no statistically meaningful evidence that the value of a licence for an umbrella brand alone was anything greater than zero.

101. On the second point, Dr Stec noted that even if a statistically significant difference between the two groups had been identified, the numerous differences between the two sets of licences meant that it would not have been possible to attribute any difference between royalty rates in the two sets to the umbrella brands alone.

102. Dr Stec made various further criticisms of Mr Wynn’s approach. I highlight the above points, however, because they formed the central basis of the cross-examination of Mr Wynn on this point. Mr Wynn’s cross-examination

103. Mr Wynn was cross-examined first (and relatively shortly) by Mr Hobbs on the general approach taken in his evidence on valuation of the notional licences. The bulk of his cross-examination was, however, conducted by Mr Hollingworth, who addressed in detail Mr Wynn’s comparables analysis, as well as his (alternative) economic benefits analysis.

104. When questioned, Mr Wynn confirmed that he did not disagree with any of Dr Stec’s calculations of statistical significance. He accepted that the differences between the sample sets used in the EY benchmarking study were not statistically significant, and that to achieve statistical significance it would be necessary to have a “rather large” dataset, which he said would have to have at least 30 data points in each set, if not more. He therefore accepted that on the sample sizes identified by EY, it would be difficult to identify a mean figure that was statistically significant. While he initially maintained that the analysis in the EY benchmarking study was a reasonable approach to follow, he ultimately accepted that the results of that study were not reliable, and that he ought to have accepted earlier that he should not place weight on that study.

105. Mr Wynn sought to fall back on the position that the figure produced by the EY benchmarking study was broadly in the same range as other licences that he had considered, such as licences for the Virgin and easy brands. His first report had, however, explicitly stated that the utility of those licences as comparables was limited since they were in different industries, and there was a lack of information on the nature of the economic relationships between the licensors and licensees. Mr Wynn ultimately accepted (albeit with considerable reluctance) that (i) there was in fact nothing other than the EY benchmarking study that produced the specific figure of 0.33% which he had used as his starting point; and (ii) if the correct starting point was not 0.33%, he had no basis to determine where between 0% and 0.33% the correct figure should lie.

106. Mr Wynn nevertheless attempted to stand by limbs (i)–(iii) of his reasoning summarised at §90 above, namely that the licences were formulated on the basis of an arm’s length principle, had real world tax implications and were subject to review by the tax authorities, and that Merck had no incentive to overstate the transfer pricing rate. None of those points, however, enables the court to place any reliance on the 0.33% figure as an appropriate starting point. As to the first point, there is no doubt that the EY analysis sought to identify an arm’s length rate, but it is now common ground that the outcome of that analysis was so unreliable as to be meaningless as a basis for such a rate. As to the second and third points, Merck’s tax incentives do not justify reliance on the 0.33% figure for the entirely unrelated purpose of the present damages inquiry (and I note that a similar argument was likewise rejected by Newey J in 32Red , §64). As Mr Wynn accepted in his cross-examination, an entirely different figure could have been selected, and he had seen transfer pricing licence rates of as low as 0.1%.

107. Mr Wynn was then asked, essentially, how his adjustments to the starting rate (which he accepted were subjective and uncertain) could be regarded as meaningful, given that he had acknowledged that the starting point of 0.33% was unreliable. He was unable to give a coherent answer to that question.

108. It is very regrettable that Mr Wynn failed to acknowledge the serious problems with the EY analysis at any point prior to his cross-examination. His abandonment of his position on the EY analysis, when pressed by Mr Hollingworth on the point, did not come as a result of any new evidence or argument, but rather reflected an inability to counter the criticisms of the EY analysis that Dr Stec had raised from the outset.

109. Having recognised, belatedly, that no weight at all could be placed on the EY benchmarking study, the only proper and objective conclusion to be drawn was that the 0.33% starting point was in consequence entirely unreliable and could not form the basis for any meaningful assessment of damages, whether adjusted or not. It is a rather obvious observation that an adjustment to a meaningless figure produces an equally meaningless figure. It is unfortunate that Mr Wynn refused to acknowledge this, instead maintaining to the end of his cross-examination that his comparables approach was preferable to an economic benefits analysis. Merck’s closing submissions

110. Merck’s written closing submissions on the comparability point maintained essentially the fallback position as set out by Mr Wynn in his cross-examination, described at §104 above. Mr Brandreth sensibly said no more about the point in his oral closing submissions.

111. For the reasons set out above, I reject that position. Merck’s comparables analysis turned on its reliance on the 0.33% figure adopted by Merck for its intragroup healthcare licences, but that figure was derived from an EY analysis which was statistically meaningless. Since Mr Wynn was unable to provide any evidence as to where the correct starting point figure should lie, as between 0% and 0.33%, if the 0.33% figure was abandoned, there is ultimately no evidential support for the adoption by the court of any specific figure at all pursuant to a comparables analysis. Any attempt to do so would be nothing more than wild speculation. Conclusion on the comparables approach

112. In light of the problems set out above, I reject Mr Wynn’s comparables analysis as a basis for the assessment of a notional licence fee in this case. Economic benefits approach The pleaded case

113. Mr Hobbs in his closing submissions pursued (albeit somewhat faintly) a submission that Merck had not pleaded a claim to damages based on an economic benefits approach, as an alternative to its reliance on a comparables analysis. This was a rather surprising submission, given that both experts considered at some length the assessment of damages based on an economic benefits approach, as an alternative to a comparables analysis; both experts agreed that in principle an economic benefits approach was a legitimate approach to adopt; and extensive submissions were made on the economic benefits analysis on both sides during both opening and closing submissions.

114. In any event, I do not accept the pleading point. Merck’s Amended Points of Claim expressly stated that Merck relied on the benefits obtained and costs avoided by MSD by the acts held to be breaches of contract and infringements of Merck’s trade marks. The avoided costs were particularised to include matters such as the costs of migrating employee email addresses from @merck.com to @msd.com, and the costs of geo-targeting of the infringing websites and developing and maintaining mirrored websites which did not use the sign Merck in an infringing way. While Merck’s pleaded quantum of loss was based on a percentage royalty approach, §26 of the Amended Points of Claim stated that: “The calculation of the precise figures claimed is dependent on information in the control of the Defendants, such as UK revenues in the relevant periods, any comparable licence figures for use of the sign MERCK by the Defendants, and accrued benefits and costs avoided by the steps taken by the Defendants. Following full disclosure and evidence, including of the Defendants’ UK revenues in the relevant periods and comparable licence royalty rates paid by the Defendants, the Claimant will provide a further calculation and reserves the right to adjust the licence figure and sum claimed as a result.”

115. In those circumstances the suggestion that an economic benefits approach to the calculation of the notional licence fee was not available to Merck on its pleaded case is hopeless. General approach

116. The experts were largely agreed as to the way in which an economic benefits approach to the calculation of the licence would work in practice. Specifically: i) From MSD’s perspective, there might be economic benefits and avoided costs which it would be willing to pay for in a hypothetical negotiation, representing MSD’s incremental gain from the notional licence relative to a situation in which MSD conducted its business without committing any acts of breach or infringement. Those benefits establish a “ceiling” for the negotiated price, in that MSD would not be willing to pay a fee exceeding its incremental gains. ii) From Merck’s perspective, the grant of a licence to MSD might harm its own business, representing Merck’s incremental loss from the notional licence. That loss establishes a “floor” for the negotiated price, in that Merck would not be willing to accept a fee below its incremental loss. iii) While the standard approach in this situation is to establish where within the “bargaining range” between floor and ceiling the notional licence fee should lie based on the bargaining power of each side (see for example my discussion at §§290–296 of Anan Kasei ), in the present case Dr Stec took a conservative approach and assumed, in Merck’s favour, that the licence fee should be set as the entirety of the incremental benefit to MSD. Merck (unsurprisingly) did not disagree with that assumption. iv) It was therefore common ground that the correct approach to a damages quantification under the economic benefits approach was to take the damages as the amount of all quantified benefits to MSD.

117. A further preliminary point concerned the date of the notional licence fee negotiation. Merck’s case was for damages for trade mark infringements from 8 March 2007 (on the basis that earlier breaches are limitation-barred under UK law), and breaches of contract from 1 January 2010 (on the basis that earlier breaches are limitation-barred under German law). Dr Stec therefore calculated damages based on the hypothetical negotiation taking place on either 8 March 2007 or 1 January 2010. MSD’s primary case, however, was that there was no basis for any claim prior to 1 January 2010, essentially since there was no evidence of any material trade mark infringement prior to that date.

118. Mr Brandreth’s oral closing submissions confirmed that if damages were assessed on an economic benefits approach, Merck did not rely on anything before 2010. He agreed, therefore, that the notional licence date should be 1 January 2010. It was common ground that the notional licence should be assumed to have covered the period up to the date of the final order of Sir Alastair Norris on liability, namely 28 July 2020.

119. On the basis of that approach, the difference between the parties’ economic benefits calculations came down to the following points: i) The avoided costs (if any) of migrating UK employee email addresses from @merck.com to @msd.com. ii) The avoided costs of geo-blocking the infringing websites in the UK and/or developing and maintaining mirrored websites without use of the Merck sign. iii) The avoided costs of developing and maintaining mirrored social media pages without use of the Merck sign. iv) The gain (if any) to MSD from web traffic diverted to MSD’s websites. v) The avoided costs (if any) to MSD of marketing the signs Merck Sharp & Dohme and MSD in the UK. vi) The avoided costs (if any) to MSD of staff training to ensure compliance with the 1970 Agreement and Merck’s trade mark rights. vii) The appropriate approach to any unquantifiable benefits. viii) The effect of inflation and discount rates on the damages calculation.

120. I will address each of these in turn. Avoided costs of email address migration

121. Dr Stec’s calculation of economic benefits included a figure attributed to the benefit that would have been obtained by MSD avoiding the cost of migrating its UK employee email addresses from @merck.com to @msd.com, if it had obtained the notional licence permitting it to continue using the @merck.com email addresses for its UK employees. He identified that cost as a one-off cost of £121,700, based on a disclosed document from MSD which purported to record the costs actually incurred by MSD in migrating its email accounts. That document was referred to in Ms Lanza’s evidence as indicating the costs of this migration exercise.

122. Unfortunately, that document was not a reliable record of the actual costs of MSD’s email migration, since it was dated 21 May 2018, whereas the evidence was that the email migration did not occur until 2020. It also appeared that the 2018 document presented a materially incomplete account of the costs of the email migration exercise. As discussed at §13 above, it became apparent during Ms Lanza’s cross-examination that this document was supplied to Ms Lanza by her solicitors, and had not been verified by her prior to her reliance on this in her witness statement.

123. In light of the unreliability of that evidence, Mr Brandreth contended that the court should rely instead on the evidence of four MSD witnesses, Adele Ambrose, Rashi Rai, Kelley Dougherty and Andrew Zager, filed in 2016 and 2018 for the purposes of earlier stages of the proceedings, and included in the trial bundle pursuant to a hearsay notice. Mr Brandreth contended that the evidence of those witnesses suggested that the actual cost of email migration was at least $610,000 upfront plus $250,000 annually, giving a total of around $3.2m or £2.4m.

124. I do not accept that submission, and did not permit Mr Brandreth to cross-examine Dr Stec on that basis. Prior to the cross-examination of Dr Stec, Merck had not ever suggested that (for the purpose of an economic benefits analysis) it would seek to argue that MSD had avoided email costs calculated by reference to the 2016 and 2018 evidence of Ms Ambrose and Ms Dougherty. Quite the contrary, Mr Wynn in his expert reports considered that it was not appropriate to include any costs of email migration, for the short reason that those were costs actually incurred by MSD, so were not avoided costs that were incremental to the notional licence. Merck’s case on damages was advanced on the basis of the evidence of Mr Wynn, with no indication prior to the cross-examination of Dr Stec that Merck was seeking to resile from Mr Wynn’s evidence and put a wholly different case in this regard.

125. The court will, of course, evaluate the quantum of damages by reference to the relevant material before it, which may well include factual as well as expert evidence. But that evaluation must necessarily be framed by the cases advanced by the parties, including in particular the evidence of the experts on which they rely. If Merck intended to argue that the evidence of Mr Wynn on this point was unsound and should be rejected by the court, and that rather than a nil figure for email migration it intended to contend for a figure of £2.4m, that point should have been identified at the latest in Merck’s skeleton argument, so that it could be explored properly with both experts, with both having had time to consider the point. It was far too late for this to emerge for the first time in Dr Stec’s cross-examination, with Mr Wynn having given his evidence already and in circumstances where (unsurprisingly) his evidence on this point had not been challenged by Mr Hollingworth.

126. The starting point in the consideration of the email migration costs must therefore be the question of whether the court should accept Mr Wynn’s evidence that the right approach should be to exclude MSD’s email migration costs (whatever they were) on the basis that these were one-off costs which MSD incurred in any event, such that those costs could not have been avoided by the grant of a licence for the period from 2010 to 2020.

127. In my judgment, Mr Wynn’s (unchallenged) explanation for his approach on this point was compelling. He was right to say that valuing the notional licence on an economic benefits approach requires an estimation of the incremental economic benefits that would have been obtained by entering into the licence. While that will include costs that would have been avoided by the agreement of the licence, a one-off email migration cost was not such a cost, because it would have been required in any event when the licence expired. Dr Stec was not specifically asked to comment on this in his cross-examination by Mr Brandreth, but it was apparent from his oral evidence that he included these costs on a conservative basis, in favour of Merck. Dr Stec did not offer any other reasoned basis for the inclusion of email migration costs in the calculation.

128. When asked about this in his closing submissions, Mr Brandreth did not disagree in principle with Mr Wynn’s analysis. He nevertheless justified the inclusion of these costs on two bases. The first was that this was a way of measuring the unquantifiable elements of the email migration process, suggesting that this was what Dr Stec intended. That is, however, not what Dr Stec said, and Mr Brandreth had not suggested this interpretation of his evidence when he cross-examined Dr Stec. It was, moreover, unclear what “unquantifiable” costs Mr Brandreth had in mind.

129. Mr Brandreth’s second response was to say that the majority of his calculated figure related to annual maintenance costs. I accept that if there was any reliable evidence of annual (rather than one-off) costs that would have been avoided by the grant of the notional licence, the avoidance of those costs could have been regarded as benefits incremental to the licence. But if there were indeed material costs attributable to email migration, there is no reason why they should not have been included in the calculations of both Dr Stec and Mr Wynn (which did include detailed figures for other annual costs, such as website maintenance costs). Mr Wynn did not, however, identify any such costs; nor did Dr Stec; and for the reasons set out above it was too late for Mr Brandreth to seek to introduce his own figures for these costs on the basis of hearsay evidence, after Mr Wynn had been cross-examined.

130. The correct approach is therefore, in my judgment, to exclude the email migration costs from the calculation of economic benefits. Avoided website costs

131. In order to address the breaches of contract and trade mark infringements, MSD implemented geo-blocking technology to prevent UK users from accessing its “Merck” branded websites. The effect of that technology was to direct users with UK IP addresses to “mirror” MSD-branded sites, which did not contain the infringing Merck branding. The costs of implementing geo-blocking were a mixture of one-off and recurring costs. The experts agreed that it was appropriate to regard some of these costs in the analysis of economic benefits. Their estimates of the appropriate costs to include were, moreover, very similar: Mr Wynn estimated £3.9m, corrected after the hearing to £3.79m, and Dr Stec estimated £3.3m (before adjustments for inflation and discounting).

132. Those differences reflected differences of approach, which were explained in detail in an Annex to Mr Wynn’s second report. The differences of approach went both ways: in some respects Dr Stec included costs which Mr Wynn did not include in his analysis; in other respects Dr Stec’s analysis assumed lower costs than Mr Wynn. It is also apparent that the expert process enabled some of the differences between the experts to be resolved, with Mr Wynn adjusting his estimates in several respects after considering Dr Stec’s analysis. Mr Wynn’s approach to this exercise was, in my judgment, conspicuously fair and objective, and I do not accept MSD’s submission that he was reluctant to make adjustments to his estimates: quite the contrary, his evidence on this part of the case demonstrates a willingness to engage with Dr Stec’s evidence and adjust his own estimates accordingly.

133. The differences between the experts fell into five main categories. It is appropriate to consider these separately.

134. One-off costs . The experts differed in their approach to one-off costs. Consistent with his approach to the email migration costs, Mr Wynn excluded various one-off costs which would have been incurred in any event, whereas Dr Stec included these. For the reasons set out above, I consider that Mr Wynn’s approach was correct.

135. Data sources . The experts used different data sources to estimate the costs that MSD avoided in relation to geo-blocking and maintaining compliant mirrored websites during the relevant period. I prefer Mr Wynn’s approach, as explained carefully in his second report, on the basis that it was more closely based on the evidence of actual costs incurred during the relevant period.

136. Attribution to UK . The experts adopted different approaches to the attribution of global costs to the UK. Dr Stec’s approach was to identify a UK apportionment of his website maintenance cost estimates, based on the UK’s share of global traffic to MSD’s websites (calculated differently for each relevant website). Mr Wynn considered that this approach was not appropriate, because it would result in the apportionment of benefits to countries in which MSD had not ever been accused of misusing the Merck sign. Mr Wynn’s preferred approach was to start off by including all of the costs required for MSD to comply in the UK, and then (in principle) to reduce that by the benefits that MSD would expect to obtain in other jurisdictions as a result of complying with the 1970 Agreement in the UK. He explained that, from an economic perspective, if the cost of compliance in the UK would have resulted in benefits associated with avoiding the need to enter into a licence in another country, that would have reduced MSD’s willingness to pay for a licence in the UK. He considered, however, that he was unable to assess the appropriate deduction because of the uncertainty of MSD’s expectations as to where it might be sued.

137. I consider that Mr Wynn’s approach is in principle the correct one, but it poses the problem that no estimate of an appropriate deduction has been given by him. This is a point on which the court therefore has to do the best it can on the available information.

138. In that regard, MSD’s closing submissions noted that Merck had brought proceedings against MSD in 13 other jurisdictions. At the time of the trial in this jurisdiction, proceedings were ongoing in both France and Germany. MSD therefore suggested that if only the benefits in France and Germany were taken into account, that would suggest an attribution of 33% to the UK on that basis. That is, in my judgment, a useful starting approach, but some allowance should also be made for the uncertain nature of the benefits in other jurisdictions. I therefore consider that it is appropriate to apportion 40% of the recurring global maintenance costs to the UK for the www.msd.com and www.msdformothers.com websites. MSD’s closing submissions accepted that no further reduction was necessary for the costs of www.msdmanuals.com, since Mr Wynn had already used a low figure (5% of global maintenance costs) for those costs, based on the evidence of MSD’s witness Mr DeFerrari. The adjusted total figure on this basis is £2.31m.

139. Merck-branded sites . Dr Stec included website maintenance costs attributable to both Merck and MSD-branded sites, whereas Mr Wynn noted that Merck would have had to maintain its Merck-branded sites in any event, such that the costs of doing so would not have been incremental to the notional licence. He therefore sought to estimate the ongoing maintenance costs specifically for the MSD-branded sites. I consider Mr Wynn’s approach to be correct.

140. Period covered . Dr Stec’s calculations assumed that the identified total recurring costs should be applied throughout the period of the notional licence, whereas Mr Wynn considered that it would be appropriate to base MSD’s avoided costs on the length of time over which MSD delayed launching each MSD-branded mirror site, so as to include only those costs which with hindsight were incremental to the notional licence. Dr Stec disagreed with Mr Wynn’s approach, on the basis that the licence fee should be modelled on the expectations at the outset of the licence.

141. On this point I prefer Dr Stec’s approach. On the basis that the notional licence must be assumed to have covered the period from 1 January 2010 to 28 July 2020, I do not consider that a negotiation at the outset of that period could plausibly have taken into account the different time points at which, in the real world, MSD introduced branded mirror sites. The reasonable approach would have been, as Dr Stec modelled, to base the negotiation on the avoided costs throughout the notional period of the licence. Applying that approach to Mr Wynn’s adjusted total (set out above) results in an uplift to the figure, producing a total of £4.33m.

142. In his closing submissions, Mr Brandreth argued that the court should adopt a figure of £5.9m for the costs of geo-blocking and development/maintenance of mirrored websites. That was of course substantially higher than the figure calculated by Mr Wynn. Mr Brandreth based that figure on his own calculations relying on (among other things) the 2016 witness statement of Ms Ambrose. To the extent that Mr Brandreth’s submissions addressed points explored in the expert evidence, I have addressed them above. I do not, however, consider it appropriate to adjust the figures discussed in the expert evidence by reference to evidence or analysis not addressed by the experts, for the reasons set out above.

143. I therefore consider that the appropriate figure to use for the avoided website costs is £4.33m (prior to any adjustment for inflation and discounting). Avoided social media costs

144. Mr Wynn included in his assessment the costs of maintaining MSD-branded social media pages, on the same basis as his inclusion of the costs of maintaining mirror MSD-branded websites. His assessment started with the costs disclosed by MSD of supporting the Merck Animal Health and MSD Animal Health social media pages, and extrapolated from those costs to calculate an estimate for the incremental costs of maintaining other MSD social media pages. His total estimate was £1.07m, taking as the relevant period for each of the Merck/MSD pairs of social media pages the period from launch of the relevant Merck-branded social media page, to the date of launch of the mirror MSD-branded page. In the case of the Merck/MSD Animal Health social media pages, for example, that gave a period from 1 April 2011 to 1 May 2017.

145. Dr Stec’s primary position was that no social media costs should be included in the calculation of economic benefits, on the basis of an absence of evidence as to the benefit from social media content, and evidence from Ms Lanza suggesting that the MSD social media costs were not separately recorded. In the alternative, Dr Stec said that if this cost were to be included, it would be necessary to consider an appropriate allocation of costs attributable to the UK, as with the avoided costs of maintaining MSD-branded websites. In closing submissions Mr Hollingworth focused on the latter point and did not vigorously pursue a case that there should be no allocation at all for social media costs.

146. I do not consider that social media costs should be disregarded entirely, as Dr Stec suggested. I have already found Ms Lanza’s evidence to be unreliable. Mr Wynn’s reports, by contrast, set out a careful analysis of the evidence supporting the inclusion of those costs, based on the costs of maintaining the Merck and MSD-branded Animal Health pages. It is reasonable for those costs to be extrapolated to other MSD-branded social-media pages.

147. Nevertheless, given my conclusions above regarding the UK attribution of the avoided website costs, it is appropriate for the same approach to be applied to avoided social media costs. An apportionment of 40% of the recurring global maintenance costs to the UK gives a revised figure of £427,426.

148. It is, however, also necessary to adjust in the other direction, on the basis of the duration point discussed at §§138–139 in relation to the maintenance of MSD-branded mirror websites. Mr Wynn’s approach was, as described above, to take as the end-date (for each Merck/MSD pair of social media pages) the date on which the mirror MSD-branded social media page was launched. Applying the analysis set out above, however, the reasonable approach would have been to base the negotiation on the avoided costs throughout the notional period of the licence. On that approach the revised figure above should be uplifted to a total of £781,703.

149. Mr Brandreth in his closing submissions (again) argued that the court should adopt a higher figure of £2m for social media costs, roughly doubling Mr Wynn’s figure. The only explanation for doing so was that Mr Wynn’s calculations were based on MSD’s disclosure, which Mr Brandreth said was incomplete. I do not accept that submission. The calculations of both experts were produced on the basis of a very detailed analysis of extensive evidence. Merck could have sought further disclosure if needed for the purposes of Mr Wynn’s report, and there was indeed considerable correspondence between the parties on the disclosure given, which eventually resolved matters by agreement. It is far too late for Mr Brandreth to seek to rely on evidential gaps in support of a submission that Mr Wynn’s figure should be uplifted by an entirely arbitrary and speculative amount.

150. I therefore consider that the appropriate figure to use for the avoided social media costs is £781,703 (prior to any adjustment for inflation and discounting). Web traffic gain

151. Mr Wynn considered there to be evidence that MSD’s infringements enabled MSD to gain website traffic at the expense of Merck, based on an analysis of SEMrush data showing traffic between June and December 2020. He relied specifically on an observation that in the period between June and December 2020, the SEMrush data showed an increase in Merck’s web traffic and a corresponding fall in MSD’s web traffic. Mr Wynn noted that the final order of Sir Alastair Norris was handed down on 28 July 2020. He considered that the shift in Merck’s and MSD’s website traffic likely resulted at least in part from changes made to MSD’s websites to comply with the July 2020 order. He quantified the benefit to MSD from increased web traffic as £1.6m.

152. The problem with Mr Wynn’s analysis arose from the evidence in the witness statement of Merck’s own witness, Mr Sielaff, regarding the use of SEMrush data. Mr Sielaff explained that SEMrush data provides an estimation of web traffic based on the search volume of a keyword and the site’s position in search results. He noted, however, that “web traffic data presented by SEMrush has a three-to-six month delay between an event occurring (like a change of content on the website) and the results of that event being visible in the SEMrush data. This is due to gradual shifts in Google keyword ranking data”.

153. On close analysis, the SEMrush data showed the most dramatic symmetrical movement in the website traffic between June–July 2020, with very little symmetrical movement from August 2020 onwards. Once the time lag of three to six months is factored in, the July 2020 order could not have been the cause of the significant shift in Merck’s and MSD’s website traffic in the SEMrush data for June to July 2020, or even the less pronounced traffic movements between July and August 2020.

154. Mr Wynn had reviewed Mr Sielaff’s evidence as to the time lag in the SEMrush data, and said that this was the basis on which he selected the period from June to December 2020. That would have been a reasonable approach if the SEMrush data showed a symmetrical shift in website traffic occurring gradually between October and December 2020, or even between July and December 2020 (on the basis that, as Mr Brandreth said, it does not follow from Mr Sielaff’s explanation that the July 2020 order would have had no impact at all on the SEMrush data until three months later). That does not, however, account for the fact that the main traffic shift identified in Mr Wynn’s chosen period occurred before the final order had even been handed down, so could not possibly have been caused by steps taken by MSD to implement the final order.

155. When cross-examined on this point by Mr Hollingworth, Mr Wynn accepted that if the flaw identified in his analysis was correct, then his figures of traffic lost and gained could not stand. He also agreed that he had not done any other analysis to try to isolate the effect of any traffic gain by MSD at the expense of Merck. For the reasons given above, I consider that Mr Wynn’s analysis was indeed flawed, and cannot be regarded as providing reliable evidence of any traffic shift from Merck to MSD. Mr Brandreth rightly pointed out that Dr Stec had not identified this particular problem with Mr Wynn’s analysis. But the problem arose inexorably from a closer scrutiny of the precise data on which Mr Wynn relied, and the Merck evidence on which Mr Wynn also specifically relied.

156. Turning to Dr Stec’s evidence, his view was that (for different reasons) there was no reliable evidence of a causal link between the July 2020 order and the pattern of the data traffic. He considered that when viewed over a longer period, the SEMrush data did not show any overall decrease in Merck’s traffic data that could be correlated to action by MSD. He also noted that using Google Analytics data presented a materially different traffic pattern over the same period. Ultimately, he said that there were many confounding factors at play in the data, and that Mr Wynn’s analysis had not controlled for the effect of those different factors.

157. It is not necessary to explore Dr Stec’s comments in this regard in any detail, given the fatal problem with Mr Wynn’s evidence discussed above. It is sufficient to say that in the absence of Mr Wynn’s analysis, there was no evidence before the court of any diversion of website traffic that could support the inclusion of this factor in the economic benefits analysis. Avoided marketing costs

158. Mr Wynn took the view that by linking its products and services to Merck in the UK, MSD may have avoided costs which it would have otherwise incurred in developing and marketing the MSD brand. While he had not seen any estimates of the cost of developing MSD as an independent brand, he considered that MSD would have had to spend at least as much as Merck’s expenditure on differentiating itself from MSD, which was one of the purposes of Merck’s rebranding exercise which commenced around 2013. Mr Wynn estimated the UK-attributable costs of the rebranding exercise at up to £2m over the period of the notional licence. On that basis his estimate of MSD’s avoided marketing costs was at least £2m.

159. Dr Stec’s position was that he had seen no evidence of any material avoided marketing costs. If, however, a portion of the costs of the Merck rebranding exercise were to be taken as a proxy for MSD’s avoided marketing costs, Dr Stec’s estimate of the appropriate figure was £566,000.

160. Starting with the question of whether any marketing figure should be included at all, Mr Wynn’s approach was in my judgment a reasonable approach to take, given the necessarily uncertain nature of the exercise. Both the trial judgment and the remitted matters judgment found that MSD’s conduct was a deliberate attempt to push the boundaries in the interests of promoting its business (trial judgment §181, remitted matters judgment §99). The remitted matters judgment commented (at §89) that MSD had implemented “a policy to bring before the market in the UK the sign ‘Merck’ in relation to the products of [MSD] at every opportunity”. It would be unreal to suggest that this did not carry a marketing advantage. In the absence of any other basis on which that advantage should be quantified, Mr Wynn’s use of Merck’s own costs in distinguishing itself from MSD was an acceptable proxy.

161. The question is then how a suitable proportion of the rebranding costs should be calculated. The experts agreed that it was necessary to start by estimating a portion of the total global costs of the rebranding exercise that could be regarded as attributable to the UK, using the UK share of MSD’s global revenues (3.2%). That produced a figure of £2m. It was also common ground between the experts that the rebranding exercise was not purely driven by a desire to differentiate Merck from MSD: there were also other factors such as geographic fragmentation and divisional sub-brands. Where the experts differed was in how to account for that in their cost estimations.

162. Dr Stec took the view that a suitable way of attributing a portion of the £2m to the purpose of differentiating between Merck and MSD was to consider the proportion of the UK figure that should be attributed to Merck’s healthcare revenues, on the basis that the overlap between Merck’s and MSD’s UK operations was in the healthcare field, and Merck’s healthcare division was therefore the division that was most likely to have been negatively impacted by MSD’s conduct. That was, in my view, a reasonable approach to take – and indeed highly conservative in Merck’s favour, given that the result of that analysis was to take all of the “healthcare” rebranding costs as the appropriate figure. That produced his figure of £566,000.

163. Mr Wynn did not put forward any alternative approach, simply saying that he did not think that there was a reliable way to work out where the right figure sat between zero and £2m. That was not a justifiable approach to take. Having recognised that taking the entirety of the UK-attributable rebranding costs would overstate the costs attributable to Merck’s differentiation of itself from MSD, it was necessary to come up with some basis to reduce the figure of £2m. Dr Stec’s approach was the only methodology available to the court which sought to do that. It may not have produced a perfect proxy, but in the context of what is inevitably a highly uncertain estimation, the court’s task is to do the best it can with the material available, rather than to strive hopelessly for absolute precision.

164. On that basis I consider that the appropriate figure to use for avoided marketing costs is £566,000 (prior to any adjustment for inflation and discounting). Avoided staff training costs

165. Mr Brandreth in his closing submissions contended that the court should include a figure of £200,000 for the avoided costs of training staff to ensure that MSD remained compliant with the 1970 Agreement. He derived that figure from the 2016 evidence of Ms Ambrose. There was no evidence on this from Mr Wynn or Dr Stec; this was therefore another attempt to increase the economic benefits figure at the end of the trial on the basis of a point not ever previously considered by the experts. It was far too late to do so, and I do not therefore consider it appropriate to include any figure for these costs. Unquantifiable benefits

166. Finally, Mr Brandreth submitted that the court should apply a broad-brush multiple or uplift to the amount quantified as benefits to MSD, to reflect the existence of substantial unquantified benefits obtained by MSD as a result of linking the sign Merck with MSD’s products and services in the UK.

167. That submission did not, however, find any support in the expert evidence. While Mr Wynn thought that the economic benefits approach would understate the true value of a notional licence, he did not offer any basis on which to uplift the figure he calculated under that approach, to reflect any further unquantifiable benefits. Dr Stec’s position was that since there was no evidence that allowed a quantification of additional benefits to MSD, that suggested that such benefits were either non-existent or immaterial. Ultimately, therefore, neither expert was able to offer any basis to uplift the economic benefits calculation to reflect any further benefits to MSD going beyond those which they had quantified.

168. In those circumstances it would be entirely inappropriate for the court to embark on a purely speculative exercise, not supported by the expert evidence, of uplifting the economic benefits calculation by an arbitrary amount. Inflation adjustment

169. The experts agreed that it is reasonable to adjust the calculated costs for inflation, so as to derive an appropriate figure assuming the date of the notional licence negotiation to have been 1 January 2010. I will leave the parties to calculate the appropriate adjustments based on the figures I have set out above. Discount rate

170. The last issue for determination is what, if any, discount is required to reflect the time value of money in relation to the payment of the licence fee at the date of the hypothetical negotiation.

171. Dr Stec’s approach was to assume that the licence fee would have been paid as a lump sum at that date, and he therefore discounted the expected cost savings back to that date using MSD’s weighted average cost of capital ( WACC ). For a 2010 negotiation date he used a WACC of 8.50%, which was MSD’s WACC at the end of 2009. Mr Wynn agreed with the principle of applying a discount rate, in order to identify a lump sum at the date of the hypothetical negotiation, but he considered that the appropriate discount rate would be a risk-free rate.

172. In his cross-examination of Dr Stec, Mr Brandreth suggested for the first time that the notional licence could have been structured as an annual payment, rather than as a lump sum at the date of the hypothetical negotiation; and contended on that basis that no discounting would be required. Again, however, I do not think that it was appropriate to introduce this point, without any foreshadowing, in the cross-examination of Dr Stec. While Dr Stec agreed that in principle a licence fee could be calculated as an annual payment, neither he nor Mr Wynn had considered in their expert reports (including the joint expert report) whether this would have been a realistic commercial option for Merck and MSD in the circumstances of this case.

173. Mr Brandreth submitted that the court should not be constrained by the views expressed by the experts when making findings about the commercial terms which could reasonably have been included in the notional licence. That might have been correct if there was a basis for such findings in the other evidence before the court. In the present case, however, the only evidence regarding the terms of the notional licence is the evidence of the two experts, Mr Wynn and Dr Stec, neither of whom considered the possibility of an annual licence fee. There is therefore no evidential basis whatsoever for a finding that the notional licence might plausibly have been structured as an annual payment rather than a lump sum.

174. In those circumstances the correct approach is to assume that the notional licence fee would have been paid as a lump sum on 1 January 2010. On that basis, as Mr Wynn agreed, it is necessary to apply some sort of discount to the inflation-adjusted cost figure. The remaining issue is what discount rate is appropriate. That breaks down into two separate questions. The first question is whether to use MSD’s WACC or a risk-free rate. The second question is what specific rate should be adopted.

175. On the first point, I agree with Mr Wynn that discounting on the basis of MSD’s WACC is not appropriate. As Mr Wynn explained, the WACC of a company is an appropriate discount rate for the overall future cash flows of a company, to take into account the market risk inherent in the calculation of those cash flows. While Mr Wynn accepted that there was some uncertainty in respect of MSD’s avoided costs, in the notional licence fee negotiation, that uncertainty was not comparable to the uncertainty of future cash flows arising from, for example, the expected future profits from a particular drug. In principle, therefore, any discount should be calculated on the basis of the appropriate risk-free rate rather than MSD’s WACC.

176. In Mr Wynn’s cross-examination, it was suggested to him by Mr Hollingworth that the risk-free rate in 2007 was about 5.25%. His position was that the discount rate should not be materially higher than the risk-free rate. On that basis, MSD submitted in its closing submissions that the minimum discount should be 5.25%. That did not, however, take into account the position (agreed at the time of the oral closing submissions) that the date of the hypothetical negotiation should be taken to be 1 January 2010. Mr Wynn was not asked about the 2010 risk-free rate in his cross-examination. In his second report, however, he commented that the Bank of England base lending rate was 5.25% at the start of the trade mark infringement period (March 2007) but fell to 0.5% by the start of the breach of contract period (January 2010). On that basis, Merck’s submission was that if a risk-free rate was adopted on the basis of Mr Wynn’s evidence, then the appropriate rate to take for a January 2010 licence fee negotiation date should be 0.5%.

177. MSD’s response, in Mr Hollingworth’s closing submissions, was that Mr Wynn’s comment related to the Bank of England base lending rate. That was, however, not the risk-free rate in January 2010. This point was then the subject of further brief written submissions following the hearing.

178. In those submissions, MSD reiterated its position that the Bank of England base lending rate was not the risk-free rate in January 2010. Rather, it said that the appropriate risk-free rate should be taken by reference to the UK 10-year Treasury bond rate in January 2010, given that the notional licence would have covered a period of 10 years. That rate was 4%. Merck’s response was that the only evidence before the court as to the risk-free rate in January 2010 was Mr Wynn’s comment regarding the Bank of England base lending rate; and that the court should therefore take that figure in the absence of any other evidence on the point.

179. MSD’s post-hearing written submission on this point is, in my judgment, the correct approach to adopt. As MSD pointed out, Merck’s position throughout the trial had been that the notional licence should be considered for the period from March 2007 to July 2020. That position was reiterated in Merck’s written closing submissions. That was the period covered by Mr Wynn’s analysis of economic benefits and was the agreed period in the experts’ joint statement. Mr Wynn’s cross-examination had proceeded on that basis, and the economic benefits figures set out in MSD’s written closing submissions were also calculated on the basis of a March 2007 licence fee date (albeit that MSD’s position remained that those calculations were conservative in favour of Merck, since it contended that there was no basis for any claim prior to 1 January 2010).

180. As set out at §116 above, Merck’s position on the start date of the notional licence only changed in Mr Brandreth’s oral closing submissions. That concession removed the dispute between the parties on this point. But the agreed licence start date must then be reflected in the quantification of the licence, including the discount rate.

181. Mr Wynn’s comment regarding the Bank of England base lending rate was clearly not intended to provide evidence about the risk-free rate in January 2010. It was, rather, simply a comment about the low interest rates prevailing at the time. There is therefore no evidence before the court as to the risk-free rate in January 2010.

182. I do not, however, consider that this is a matter that requires expert evidence. The risk-free rate is the expected return demanded by investors on an investment with no risk of loss of the principal. The standard way of assessing that is to use, as a proxy, the yield on the bond issued by an AAA-rated government of the same currency as the investment, over the same period of time as the investment. In the present case, an appropriate assessment of the risk-free rate in January 2010 was therefore, as MSD submitted, to look at the 10-year UK Treasury bond rate at that time. Merck conspicuously did not offer any other basis for the identification of the risk-free rate, other than its reliance on the 0.5% figure in Mr Wynn’s evidence, which I have rejected.

183. The appropriate discount rate to use is therefore, in my judgment, the 4% risk-free rate identified by MSD. Conclusions

184. For the reasons set out above, my conclusions are as follows: i) It is appropriate to award damages in the present case based on a notional licence fee. ii) I do not consider that Merck’s comparables analysis is a reliable basis for assessing the value of that notional licence. iii) The appropriate basis for the valuation of the notional licence is therefore the economic benefits approach. That is an approach open to Merck on its pleaded case. iv) Applying an economic benefits approach, the notional licence fee should be calculated on the following basis: a) No allowance should be made for avoided email migration costs, diversion of web traffic or staff training costs. b) Prior to adjustments for inflation and discounting, the notional licence fee should comprise avoided website costs of £4.33m, avoided social media costs of £781,703, and avoided marketing costs of £566,000. c) Those figures should be adjusted for inflation and discounting, using a discount rate of 4%.

185. I will hear further submissions on the precise licence figure on the basis of the conclusions set out above.