[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales High Court (Commercial Court) Decisions
You are here: BAILII >> Databases >> England and Wales High Court (Commercial Court) Decisions >> Bikam OOD & Anor v Adria Cable SARL [2013] EWHC 1985 (Comm) (12 July 2013) URL: http://www.bailii.org/ew/cases/EWHC/Comm/2013/1985.html Cite as: [2013] EWHC 1985 (Comm)

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

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
COMMERCIAL COURT

		Royal Courts of Justice 7 Rolls Building, Fetter Lane London, EC4A 1NL
		12/07/2013

The Hon. Mr Justice Popplewell :

Introduction

1. By a Sale and Purchase Agreement dated 7 October 2009 ("the 2009 SPA"), the Defendant ("the Buyer"), purchased the majority shareholding of a Bulgarian satellite television company from the Claimants ("the Sellers") for a total price of €6,450,000. The price was payable in two tranches, €850,000 being payable (and paid) on completion and the balance at a later date contingent on various events. Under clause 3 of the 2009 SPA, the price potentially fell to be adjusted following a review of the net debt of the company by PricewaterhouseCooper ("PwC") which was to take place following completion. Following PwC's review, the parties agreed that the net debt adjustment should reduce the second tranche of the price by €1,406,000 to €4,194,000. Just over one year later, the Buyer sold its shareholding, and further shareholdings which it had also acquired, to a third party for less than €1. This sale triggered the Buyer's obligation to pay the second tranche of the price.
2. In this action, the Sellers bring a claim for the second tranche of the purchase price in the sum of €4,194,000. The Buyer defends the claim for the price, and counterclaims, on the basis of a number of alleged breaches of warranty. The main allegation is of breach by the Sellers of a warranty that the company had 80,000 "DTH subscribers" on completion. In addition to their claim for the second tranche of the price, the Sellers seek to undo the effect of the agreed net debt adjustment and claim back the €1,406,000 reduction in the price.
3. The company was called Interactive Technologies AD ("the Company"). It was founded in 2001 by two businesswomen, Ms Dimitrova and Ms Ivanova. It provided digital direct-to-home (DTH) satellite television services. It purchased television (and radio) content from various companies and broadcast the content via satellite directly to subscribers' homes. Although the Company had some other sources of income, its key revenue stream came from fees paid by subscribers, which amounted to between 70% and 80% of the Company's total revenue.
4. Prior to the sale, the First Claimant held a 13% shareholding and the Second Claimant a 61% shareholding in the Company. Ms Dimitrova is the majority shareholder of the First Claimant. Ms Ivanova is the majority shareholder of the Second Claimant. Prior to the sale, Ms Dimitrova was the Company's CEO. Its Commercial Director was Mr Sherbanov, who is Ms Dimitrova's brother. Its Marketing Manager was Mr Kanchev, who is Ms Dimitrova's son. Ms Ivanova worked at the Company, without any formal title, in the same office as Ms Dimitrova. All ceased to work for the Company immediately following the sale.
5. The Buyer is owned by Mid Europa Partners ("Mid Europa"), a private equity firm focused on the markets of Central and Eastern Europe. On 27 June 2007 Mid Europa had acquired a substantial controlling stake in Srpske kablovske mreze d.o.o., translated as Serbian Broadband ("SBB"), a leading cable and satellite television and internet service provider in Serbia.
6. The Company had been built up from scratch into one of the leading DTH satellite companies in Bulgaria, competing successfully at the higher end of the market. However, cash-flow difficulties and the credit crunch forced Ms. Dimitrova and Ms. Ivanova to find a buyer for the Company. They chose to sell the Sellers' shareholding to SBB in 2008, but the purchase fell through when SBB failed to obtain the necessary financing. However, Mid Europa remained interested in the Company and the Buyer eventually purchased it by the 2009 SPA on 7 October 2009.
7. The Company's DTH business operated as follows. Subscribers paid for a service consisting of individual conditional access to radio and television programs via digital satellite transmission. Subscriptions were typically monthly in advance, but were sometimes for longer periods as a result of promotions. In order to access the Company's content, subscribers required a smartcard to insert in a set-top box in their home which received the downlinked signal from the satellite. The subscriber did not need to use a set-top box provided by the Company, and the smartcards were not tied to a particular set-top box, address or even country; it was only necessary to have a set-top box which was compatible with the Company's encryption software.
8. Access to content was controlled by a conditional access system ("CAS"), which ensured that only those subscribers who had paid for the service could access the content. Subscribers had to pay at least monthly in advance. If they failed to pay in advance, their smartcards would be electronically de-activated in the CAS. Upon subsequent payment, the subscriber's smartcard would be re-activated. It was not uncommon for subscribers to fail to pay at the end of one month but do so only in the early part of the following month when they next wanted to watch the Company's content. It was therefore a feature of the Company's subscription base that at the beginning of the month a proportion of subscribers would be deactivated, who would then become reactivated over the early part of the month by renewing their monthly subscription. After 90 days in default of payment, subscriptions were cancelled. Therefore the Company recorded each subscription as "active", "deactivated" or "cancelled".
9. The Company contracted with television content providers for the supply of television channels, which the Company then packaged and sold to its subscribers. The Company offered a variety of economy and premium packages with different channel options, and paid the television content providers, usually on the basis of how many subscribers received that channel. A subscriber could take out more than one subscription, in which case the additional subscription would have an additional smartcard and its own channel content. The subscriptions would be billed, activated and de-activated independently of each other and could be used by different people, for example other family members, on different televisions.
10. Every package offered by the Company included the "economy" package of television channels, which was the minimum content for every subscription and came with every smartcard. There were a variety of different packages which included additional combinations of channels or "add-on" themes. To purchase such additional content, it was always necessary to have purchased a subscription to at least the lowest package of the Company; subscribers could not subscribe only to premium content.
11. The Company's subscribers paid their subscriptions through distributors, the Company's own shops and installers, and to a lesser extent by electronic payments. As at September 2009, 5% paid electronically, with cash payments to distributors and the Company's own sales network accounting for the remainder.
12. Separately from its customer subscription service, the Company also sold "uplink and transmission" services to television content providers. The Company acted as Eutelsat's distributor in the region. The Company would uplink content, which cable networks then redistributed via their cable lines to their own subscribers. The Company also offered subtitling and dubbing services. The Company charged television content providers for these uplinking services, bringing in revenue of about BGN 3.5m in 2009.
13. One category of subscription at the heart of the warranty dispute was a special promotional "package" called "Second Television". I have put "package" in inverted commas because it is controversial whether it should properly be descried as a package in the same sense as other packages offered by the Company. This "package", whose terms varied from time to time depending on the promotion, comprised a second subscription, available to existing subscribers only, at a cheaper price than the first subscription. The Second Television subscription came with a separate smartcard to which the subscriber could add premium content and which could be used independently and, I conclude on the evidence, in a separate location, from the first smartcard. Individuals who wanted more than one subscription could (and did) purchase additional first subscriptions to any package of the Company; they did not have to purchase the Second Television subscription as their second (or third etc.) subscription. Because of the terms of the promotions, however, the Second Television subscription proved popular.

Narrative

14. In 2008 the First and Second Claimants held 74% of the shares in the Company. The remaining 26% were held by Maxstone Services Ltd ("Maxstone"). In the early summer of 2008, Maxstone agreed to sell its shareholding to CableTEL. CableTEL then approached the Sellers with an offer of €40m for their 74% interest in the Company. In June 2008 negotiations started between SBB and the Sellers for SBB to purchase the Company. SBB had become aware of the potential CableTEL deal and persuaded the Sellers that SBB was a more suitable buyer because of its ready access to finance and strategic view for the Company's future. SBB wished to acquire 100% of the Company, and therefore asked the Sellers to intercede on their behalf with Maxstone to cancel the sale with CableTEL, which the Sellers did.
15. SBB and PwC conducted due diligence on the Company in July and August 2008. As part of this process, the Sellers and SBB exchanged business plans and other documents, some of which I address in more detail below. As part of the due diligence Ms Vasiljevic, SBB's chief operating officer, spent about 10 days at the Company's offices in Sofia. Following due diligence, SBB and the Sellers entered into an SPA on 15 September 2008 ("the 2008 SPA"), by which SBB agreed to purchase the Sellers' shares for €45m (plus up to €5m more dependent upon the Company's earnings for 2008).
16. SBB failed to obtain finance and the parties agreed to extend the longstop date for payment to 31 January 2009. Further due diligence took place during this extended period, but SBB was unable to obtain the necessary finance and the 2008 SPA was cancelled.
17. Following the collapse of the 2008 SPA, Mr Knorr of Mid Europa approached the Sellers in April 2009. There followed negotiations between the Sellers and Mid Europa in June, July and August 2009. In August 2009 PwC and SBB/the Buyer embarked on a further round of due diligence on the Company. As part of SBB's due diligence on behalf of the Buyer, Ms Vasiljevic visited the Company's offices on 26 and 27 August 2009.
18. A draft SPA was circulated by Mr Tzvetkov, a director of Mid Europa, on 7 September 2009. Negotiations on this draft continued in September and into October 2009 and were concluded with the signing of the agreement on 7 October 2009.
19. The 2009 SPA included a warranty by the Sellers that the Company's net debt was BGN 42,875,990. Pursuant to clause 3, the net debt of the Company was to be the subject of a review following completion, and the price fell to be adjusted if the difference between the warranted and actual net debt figures was more than €100,000. The parties were to endeavour to agree the net debt following a review by PwC; failing agreement, another accountancy firm would be appointed at the Sellers' cost to arbitrate the issue. As set out in more detail below, there followed a period of review by PwC which largely excluded the Sellers and during which, the Sellers say, the Buyer failed to provide relevant information to PwC and/or failed to allow the Sellers access to that information. PwC produced a draft net debt report to the Buyer on 16 November 2009. PwC's net debt statement was sent to the Sellers on 1 December 2009 and recorded a net debt of BGN 50,899,000 a difference of BGN 8,022,000 (more than €4m).
20. The Sellers were surprised by this figure and sought access to the relevant accounting information and documents on which the net debt review was carried out. The Sellers say that they were not granted full and proper access in response to this request, which is denied by the Buyer.
21. There were a series of meetings on 11, 14, 15, 16 and 19 December 2009 variously between the parties and PwC. The upshot was an agreement recorded in a written Net Debt Agreement signed and dated 28 December 2009. By this agreement the parties agreed a compromise reduction of BGN 2.75m (€1,406,000) in the price, to account for the Buyer's claim that the net debt was significantly higher than warranted net debt. The effect was to reduce the second tranche of the price to €4,194,000. The Sellers' case is that the Net Debt Agreement was entered into following misrepresentations by the Buyer about the accuracy and basis of the net debt, and breaches of the 2009 SPA in failing to provide to PwC and the Buyer access to relevant information and documents.
22. By mid-2010 the Company was in serious financial difficulty. A meeting between the Buyer's lawyers and the Company's main creditor, Bulbank, was held on 12 August 2010 at which it was agreed that the Company would not file for bankruptcy until the end of October 2010 to enable the bank to find a buyer for the Company. Newgrant Development Ltd ("Newgrant") expressed an interest in purchasing the Company's assets, although it was reluctant to purchase the shares. Newgrant agreed to buy the Company's shares for nominal consideration on 9 November 2010. This triggered the Buyer's obligation to pay the second tranche of the purchase price under clause 3.16 of the 2009 SPA within 30 days of 9 November 2010, that is 8 December 2010.
23. The Buyer's lawyers wrote to the Sellers on 9 November 2010 asserting for the first time that "there have been very serious breaches of the Sellers' Warranties" causing loss exceeding €10.7m. No further explanation of the alleged breaches was given until 8 December 2010, when the Buyer's lawyers wrote to inform the Sellers of the sale to Newgrant and to assert that the second tranche of the purchase price was not due because of 16 distinct breaches of warranty by the Sellers. Of those, the Buyer maintained five alleged breaches at trial. The main one is an alleged breach of the warranty that the Company had 80,000 "DTH Subscribers" on completion.

DTH Subscriber Warranty

24. By clause 7 of the 2009 SPA the Buyers warranted the truth of the Warranties in Schedule 2 at 7 October 2009. Paragraph 20 of Schedule 2 contained the DTH Subscriber Warranty in the following terms:

"20. DTH Subscribers

(a) For the purposes of this Paragraph 20, "DTH Subscribers" means subscribers who have contracted, directly or indirectly, with the Company to subscribe to the lowest package of the Company (not including premium channels and HBO) and who are not 90 days or more in default of payments due and payable under such contract.

(b) As at the date hereof, there are a total of at least 80,000 DTH Subscribers."

25. The Buyer alleges that, contrary to the DTH Subscriber Warranty, the Company only had 65,394 DTH Subscribers at completion. The Sellers contend that there were 82,569 DTH Subscribers.
26. There are two aspects to the DTH Subscriber Warranty dispute. The first is whether and to what extent certain categories come within the definition of "DTH Subscribers", namely Second Television, hotel, and cable subscriptions. Of these the most important is Second Television subscriptions, which account, on the Buyer's figures, for 8,468 subscriptions. The second aspect is a dispute over the number of all types of subscribers at 7 October 2009, which arises because the Sellers' figures are based on data downloaded from the Company's systems on 7 October 2009 by Mr Sherbanov, whereas the Buyer's figures are based on data produced and preserved by a system back up before the Company was sold to Newgrant in late 2010. The rival figures differ substantially and are not reconcilable. Each side argues for the reliability of its data and the unreliability of the other side's data.

Do Second Television subscriptions qualify as DTH Subscribers?

27. This a question which falls to be determined as a matter of construction of the DTH Subscriber Warranty. The approach to construction has recently been authoritatively restated in Investors Compensation Scheme v West Bromwich Building Society [1998] 1 WLR 896 and Chartbrook v Persimmon Homes [2009] 1 AC 1101. In Investors Compensation Scheme Lord Hoffmann said at 912H that:

"Interpretation is the ascertainment of the meaning which the document would convey to a reasonable person having all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contact."

In the latter case Lord Hoffmann said at [14]:

"….the question is what a reasonable person having all the background knowledge which would have been available to the parties would have understood them to be using the language in the contract to mean."

28. The background knowledge which is relevant is not confined to facts and matters actually known to both parties; it extends to facts and matters which a party could reasonably expect the other to know or find out about if he wanted to.

This is the knowledge "reasonably available to the parties" in Lord Hoffmann's formulation.

29. The packages offered by the Company were set out in two documents of relevance which were known to the parties. These were Item 18.2 of Schedule 4 to the 2008 SPA, which became Item 18.2 of Schedule 17 of the 2009 SPA, which was a due diligence document made available to the Buyer entitled "A description of all services and/or electronic communications networks provided by the Company"; and the Company's website at about the date of the 2009 SPA, which listed the various package subscriptions. Both of these documents show that there were two categories of content being offered by the Company:

(1) The first category comprised the following packages: Economy (BGN 8.6); Practical (BGN 13.80); Basic (BGN 17.80); Special Economy (BGN 13.60); Special Dynamic (BGN 19.80); Special Practical (BGN 19.80); and Premium Package (BGN 26.80). All of these packages had a subscription to the Economy package channels, to which further basic content (i.e. the topics/themes) and/or premium content was added at increased cost to the subscriber for all packages except Economy.

(2) The second category comprised the following special packages: HBO; HBO/Cinemax; X-Full and HD. These packages were premium content only packages and did not include a subscription to the Economy package.

30. Mr Sherbanov's evidence, which I accept, was that the premium packages were not available as an independent subscription without a person having subscribed to one of the packages including the Economy package element.
31. The Second Television promotions were not referred to as a package in either document.
32. Mr Sherbanov described the way in which Second Television subscriptions worked and their rationale. I found him to be an honest and impressive witness. What he told me was knowledge reasonably available to the Buyer, who spent many days doing due diligence on the Company with an opportunity to consider the make up of its subscriptions. On any view Second Television subscriptions accounted for a significant proportion of the subscriptions (of the order of 10%) and the Sellers could reasonably have expected the Buyer to become familiar with them to the extent that subscriber numbers were of importance to the Buyer. Mr Sherbanov's evidence included the following:

(1) According to official figures at the National Statistic Institute, 62% of Bulgarian families own a second home. The Company's existing customer base was therefore regarded as a potential market for new subscriptions. Because smartcards were not tied to any specific location, it was a common practice for people to buy subscriptions for friends or family as a gift, for example for children who lived and studied away from home, or family members who lived in regional parts of Bulgaria or abroad. The Company was keen to tap into this market and ensure that its existing subscribers chose to purchase any new, additional subscriptions from the Company. A number of loyalty and retention promotions were introduced, but the most popular proved to be the Second Television subscription.

(2) In order to qualify for the Second Television subscription, the subscriber had to have an existing subscription to one of the Company's packages at the time of purchase. The Company's distributors, or other sales agents, would check the individual or company ID number against the subscription records to see whether there was an existing smartcard recorded against that number. If this was confirmed, then the sale of the Second Television promotional subscription would be approved, and a further separate smartcard would be authorised.

(3) Once a Second Television subscription was purchased, the two (or more) subscriptions were not linked in the system. Every Second Television subscriber had to receive a unique smartcard in order to provide access to the specific content of that subscription. This meant that his two subscriptions operated independently. The subscriptions could, and did, contain different channel content, and subscribers could, and did, activate and deactivate their subscriptions independently. If the first subscription lapsed and the subscriber continued to pay for the second, only the first was deactivated; the other remained active and continued.

(4) The price and content of the Second Television subscription (like any other package) depended on the promotion running at the time. Second Television promotions were usually structured to mirror the content in the Economy package or the Basic package. The Second Television subscriptions were usually cheaper than a first subscription, because the Second Television promotion was designed to encourage loyalty. But it was not necessary for a subscriber to buy one of the Second Television packages if he wanted to buy an additional subscription from the Company. A subscriber could choose to buy any other package available, and many did.

(5) Second Television subscribers, like subscribers to any package, could choose to switch to a different package of content on their Second Television subscription, in accordance with the rates currently offered by the Company for the different packages of content. A subscriber with a Second Television subscription for use at his holiday home might, for example, choose to upgrade the holiday home subscription to a Premium package just for the holiday months and later downgrade.

33. There was a dispute about whether Mr Sherbanov was right in saying that the Second Television subscription was permitted to be used at a different address or in a different "household" from the first subscription, and in particular whether such use was consistent with some of the due diligence documents and what Ms Vasiljevic said in cross examination that she had been told during her due diligence in 2008. This dispute has no significant bearing on my approach to construction of the warranty. Had I regarded it as material, I would have rejected Ms Vasiljevic's evidence and found that Mr Sherbanov's description of the practice was not inconsistent with the documents relied upon by the Buyer.
34. Mr Sherbanov was also challenged on his evidence that the Second Television subscription did not give access to the same package as the customer's first subscription. The promotion being run on the website of one of the distributors, Eco Engineering, suggested that for the Second Television subscription of BGN 7, the subscriber was entitled to the same package as his primary subscription but without such premium channels as that might include. Mr Sherbanov's evidence was that that was simply the form of the promotion being run by that distributor at that time. The Company's website suggested that at around the time of the 2009 SPA the promotion being run by the Company was to similar effect; that reconstruction of the website was only adduced in evidence by the Buyer after Mr Sherbanov had concluded his evidence and he was not asked about it. It may not have reflected the practice with promotions at other times. I am prepared to assume in the Buyer's favour that the predominant practice with Second Television subscriptions was that they could include the full package of the first subscription but without the premium content. Nevertheless the two subscriptions might have different content. A Second Television subscription was amenable to the subscriber choosing to add additional content, including but not limited to premium content, in just the same way as a first subscription.
35. The evidence from the Sellers' witnesses, which I accept, was that the practice of the Company, when reporting subscriber numbers to content providers for the purposes of payment for content, was that Second Television subscriptions were included as separate subscribers. Ms Vasiljevic accepted that her colleagues' due diligence included familiarisation with the methods of reporting to content providers, although she denied that she was personally told that Second Television subscribers were separately reported. Nevertheless it is clear that the practice of treating Second Television subscriptions as separate subscribers for the purposes of reporting to, and paying, content providers was one which was continued by the Company under the ownership of the Buyer. This was apparent from an email from Commplus to Ms Vasiljevic dated 28 January 2010 and a spreadsheet in almost identical form used in April 2010 to report subscriber numbers to content providers.
36. Considerable time was spent in the evidence on an analysis of how Second Television subscribers were dealt with in business plans and spreadsheets prepared by the parties and exchanged during 2008 and 2009, to which I must turn. I remind myself that these are only a legitimate aid to construction insofar as they were shared between the parties, although purely internal documents may cast light on what was communicated orally between the parties.
37. On 8 August 2008 Ms Dimitrova sent an email to SBB attaching a letter and three attachments referred to in the letter. The letter addressed average revenue per unit ("ARPU") and for those purposes was supported by an average price analysis spreadsheet ("the 0808 average price spreadsheet") and an ARPU analysis spreadsheet ("the 0808 ARPU spreadsheet"). The letter also addressed uplink revenues, programme costs and churn, and concluded that the EBITDA projected for 2008 would be €4,963,000. These were supported by an Excel spreadsheet attachment entitled "BPL0808withcorrections spreadsheet". This last comprised a number of individual spreadsheets with a wealth of data, including those to which I will refer by their tab references as 0808 "Key Performance", "P&L DTH", "DTH Subscribers", "DTH Subscribers Total" and "M-S". So far as relevant these documents included the following information:

(1) The 0808 average price spreadsheet used subscriber numbers which it is now common ground did not include Second Television subscriptions, but that would not have been apparent from the document itself.

(2) The 0808 ARPU spreadsheet gave a breakdown of the numbers of subscribers and revenue for each of the first six months of 2008. In doing so it listed separately the figures for package subscriptions and for "Second TV" subscriptions and calculated the ARPU as an average of the total subscription numbers and revenues in both categories. In other words it treated Second Television subscriptions as subscribers for the purposes of calculating the average revenue per subscriber. This was obvious from the face of the document.

(3) The 0808 Key Performance spreadsheet, which as its name suggests was a summary of data drawn from the other spreadsheets, gave a number of subscribers for each of the calendar years 2007 (actual), and 2008 to 2010 (projected). An analysis of the source of this data from the function box and the underlying DTH Subscribers, DTH Subscribers Total and M-S spreadsheets reveals that the subscriber totals treated Second Television subscriptions as separate and additional subscribers making up these totals.

(4) The 0808 DTH Subscribers Total spreadsheet counted and included Second Television subscriptions as "DTH Subscribers".

(5) In the 0808 P&L DTH spreadsheet turnover from individual subscriptions was identified and listed in a separate line from "Turnover second receiver tax". The ARPU was calculated on revenue which included revenue from both categories.

(6) The 0808 M-S spreadsheet has titles to rows for subscriber numbers which include a row for DTH basic subscribers and another row for "second receiver subscribers". This treats Second Television subscriptions as "subscribers". In the rows dealing with revenue it again distinguishes between fees for the basic package, fees for premium options and "second receiver tax", the last being the revenue from Second Television subscriptions.

38. On 18 August 2008 Ms Boklag, the deputy CEO of SBB, sent an email to Ms Dimitrova attaching two spreadsheets. One was a Revenue and ARPU analysis seeking to identify the main differences in the assumptions between the parties ("the SBB Revenue and ARPU Analysis spreadsheet"). The other was a projected income statement for the Company which included a Revenues spreadsheet ("the SBB Revenues spreadsheet"). These included the following information:

(1) The SBB Revenue and ARPU Analysis spreadsheet split out ARPU by reference to package subscriptions and the additional revenues (including those from Second Television subscriptions) separately. It sought to calculate the projected penetration rates of premium and Second Television subscriptions as an uplift based on the number of basic subscriptions.

(2) The SBB Revenues spreadsheet separated out basic service subscriptions, premium service subscriptions and "second receiver" subscriptions.

39. The 2008 SPA, concluded on 15 September 2008, contained the DTH subscriber warranty, at paragraph 20 of Schedule 2, in the same terms as in the 2009 SPA.
40. As part of the due diligence in 2009 Ms Vasiljevic visited the Company's offices in Sofia on 26 and 27 August 2009. Two spreadsheets of potential significance produced at this time were a subs_and_revenues_2009.xls spreadsheet and a churn.xls spreadsheet. There is a dispute as to how and by whom they were produced, which is of potential importance because they treated the Second Television subscriptions as separate and additional active subscribers, in the following way:

(1) The copy of the subs_and_revenues_2009 spreadsheet in the form disclosed by the Sellers has 279 rows, in Bulgarian. It listed all the Company's 277 offerings, under the heading "promo/pack". Column B was headed TP meaning type of package, in which against each offering there was a "1" or a "0". All those which were separate subscriptions which required a separate smartcard were flagged with a "1" in column B; all "add-on" offerings were flagged with a "0" in column B. The Second Television subscriptions, of which there were at least 12 different variations, were all marked with a "1". This made it clear to the Buyer that each subscriber to those promotions/packages was counted as a separate subscriber by the Company and reported as such to the content providers. This was something which was explained to Ms Vasiljevic and, in any event, would have been clear from her review of that spreadsheet.

(2) There is a different version of the subs_and_revenues_2009 spreadsheet with additional figures in rows 282 to 290. These additional rows were in English. These include row 286 headed "Total active subscribers". The total active subscribers total in row 286 included, as separate and additional subscribers, Second Television subscriptions, being all those flagged with a 1 in column B. Some of the figures in the additional rows were not derived from the figures earlier in the table. It was common ground that this information must have come from the Sellers, although it does not follow that it was input into the spreadsheet by someone on the Sellers' side rather than provided by the Sellers and imported into the spreadsheet by Ms Vasilijevic.

(3) The churn spreadsheet had headings in English. It calculated the number of active subscribers based on the number including Second Television subscriptions rather than the number excluding Second Television subscriptions. Although for the months from January 2009 it listed in a separate column "2^nd TV" subscriptions, they were included as additional subscriptions in the active subscriber column. Therefore whoever prepared this spreadsheet must have considered active subscribers to include Second Television subscriptions, and it was obvious from the face of the document that it was prepared on this basis.

41. Ms Vasiljevic's evidence was that Mr Kanchev provided her with the subs_and_revenues_2009 spreadsheet on a memory stick in the form which had the additional rows at 282 to 290. Ms Vasiljevic's evidence was that when she realised that the figure included Second Television subscriptions, she expressly made it clear to Ms Dimitrova and Mr Kanchev that additional payments for the Second Television offers did not count as separate subscribers. She requested that Ms Dimitrova prepare a separate table which listed the number of second set-top boxes in it. Her evidence was that Mr Kanchev then provided her with the churn spreadsheet, again on a memory stick.
42. The evidence of Ms Dimitrova and Mr Kanchev was that Mr Kanchev had emailed Ms Dimitrova a copy of the subs_and_revenues_2009 spreadsheet. This was the 279 row spreadsheet, in the form disclosed by the Sellers, which did not include the "Total active subscribers" total. Ms Dimitrova's evidence was that she saved this to her computer and then sent it on to Ms Vasiljevic. Ms Dimitrova and Mr Kanchev both deny performing the calculations in rows 282 to 290, deny having provided the churn spreadsheet and deny having been told by Ms Vasiljevic that Second Television subscriptions should not be counted as subscribers.
43. I prefer the evidence of the Sellers' witnesses on this dispute about the 2009 spreadsheets for the following principal reasons:

(1) Ms Vasiljevic prepared a further version of the subs_and_revenues_2009 spreadsheet, which it is common ground was not sent to the Sellers. This was the subs_and_revenues_2009_vv" spreadsheet. She made a correction to the function formula to correct an error in the subs_and_revenues_2009 calculations in row 286. This then gave a total subscriber number which included Second Television subscriptions. Her ARPU calculation was also based on the total number of subscribers including Second Television subscriptions. I found her explanation in cross examination that this was only done to correct the data as presented by the Sellers implausible and unconvincing. This was her working document in which she was seeking to analyse the data.

(2) In the churn spreadsheet the "active" column contains the figures taken from the subs_and_revenues_2009_vv spreadsheet. The "active" subscribers must have been taken from this spreadsheet, rather than subs_and_revenues_2009 because the figures are those following Ms Vasiljevic's correction of the formula error in subs_and_revenues_2009. This shows that the churn spreadsheet was produced after subs_and_revenues_2009_vv and incorporated data from that spreadsheet. The person who prepared the churn spreadsheet must therefore have had access to Ms Vasiljevic's amended spreadsheet. This must have been her, not Mr Kanchev or Ms Dimitrova.

(3) Ms Vasiljevic's evidence on this area of dispute was not consistent. In her witness statement and initially in cross examination she stated that she was given both spreadsheets by Mr Kanchev on a USB stick in a meeting. That cannot be right because, as set out above, the churn spreadsheet was prepared using data from subs_and_revenues_2009_vv, a spreadsheet which Ms Vasiljevic prepared using the subs_and_revenues_2009 document. Ms Vasiljevic also stated in her first witness statement that she asked for the churn spreadsheet to check her calculations in subs_and_revenues_2009_vv, which is inconsistent with her statement that she received both at the same time. When this inconsistency was put to Ms Vasiljevic, she gave a new explanation that she was given the two spreadsheets a few hours apart.

(4) It was unclear from Ms Vasiljevic's witness statement whether she was saying that the alleged conversation, in which she told Ms Dimitrova and Mr Kanchev that additional payments for the Second Television offer did not count as separate subscribers, took place during the 2008 due diligence or that in 2009. She ultimately said that it had happened on both occasions. But no sensible conversation about what counted as a "DTH Subscriber" could have taken place during the due diligence in 2008 because the first draft SPA had not even been circulated by this point. In any event Ms Vasiljevic was not involved in the negotiation of the SPA, and there was no reason in 2008 or 2009 for her to have attempted to agree what fell within and outside the contractual definition.

(5) Ms Vasiljevic was involved in supervising the running of the Company in the months following completion. Had she thought that the subscriber warranty did not include Second Television subscriptions, it would have been readily apparent soon after completion that there was a significant shortfall. Irrespective of who is right about the reliability of the rival data sets now before the court, the exclusion of Second Television subscriptions from the number of subscribers shown in the data available at the time would have made it obvious that there were well below 80,000, which it was Ms Vasiljevic's evidence was a figure of importance to her in giving the Company critical mass. Yet she did not make any complaint at the time, either within SBB or to the Sellers, about a shortfall in subscriber numbers. This conduct is only consistent with her understanding at the time that Second Television subscriptions were counted as subscribers. At one point in her evidence she said that she had asked the IT manager of the Company at the end of October or beginning of November 2009 how many active subscribers there were, and been told about 62,000. I found that incredible. She could not offer any credible explanation as to why that had not generated a complaint of breach of the warranty or any internal correspondence or concern.

(6) By contrast with Ms Vasiljevic, I found Ms Dimitrova to be a straightforward and persuasive witness on this subject. It is difficult to see how the Sellers' 279 row version of the subs_and_revenues_2009 could have been created and disclosed, consistently with Ms Vasiljevic's version of events, without some deliberate dishonesty on Ms Dimitrova's part. It is improbable that she saved a draft version and not a final version on her computer. No such dishonesty (or indeed any other explanation for the allegedly incomplete document) was alleged against Ms Dimitrova and I would unhesitatingly reject such a suggestion.

(7) I have not overlooked the fact that the day after Ms Vasiljevic completed her due diligence, she reported a summary of the Company for the first 6 months of 2009 to Mr Solak. In this summary, she explained that "The total number of active users is 75,000 of whom 9,500 have a connection for a second TV". As is clear from subs_and_revenues_2009_vv.xls, which had been produced the day before, the 75,000 figure does not include Second Television subscriptions. But I do not regard the fact that Ms Vasiljevic was reporting in these terms as something which lends significant support to her version of events. Nor do I attach significance to the fact that the additional rows in subs_and_revenues_2009 and the headings in the churn spreadsheet were in English. Ms Vasiljevic would have prepared anything that she intended to show to the Sellers or to Mr Tzvetkov in English.

44. Accordingly I find that as part of the due diligence in 2009:

(1) The Sellers presented the data about the Company's subscribers and subscriptions to Ms Vasiljevic on the basis that Second Television subscriptions were treated and counted as separate and additional subscribers.

(2) Ms Vasiljevic knew that the Sellers were treating Second Television subscriptions in that way.

(3) Ms Vasiljevic said and did nothing to suggest that she was treating Second Television subscriptions differently. On the contrary, in the documents she prepared as part of the internal analysis on behalf of the Buyer, she herself treated Second Television subscriptions as counting towards the total number of subscribers.

45. Against this background of knowledge reasonably available to the parties I turn to the construction of the DTH Subscriber Warranty in paragraph 20. Shorn of its context, one might have started with an inclination to think that the reference to a "subscriber" was a reference to a person. This was the Buyer's submission as to how the warranty fell to be construed. But the context shows this to be mistaken, and that "subscribers" is a measure of the number of subscriptions, not of persons. It is clear from the spreadsheets exchanged in both 2008 and 2009 that both parties treated Second Television subscriptions as capable of being described as "subscribers" in their own right. The 0808 ARPU spreadsheet treated Second Television subscriptions as separate subscribers for the purposes of calculating the average revenue per subscriber. In the 0808 Key Performance spreadsheet, Second Television subscriptions are treated as separate subscribers making up the total number of subscribers. The 0808 M-S spreadsheet treats Second Television subscriptions as "subscribers". The subs_and_revenues_2009 spreadsheet and churn spreadsheet treated the Second Television subscriptions as separate active subscribers.
46. Moreover, Mr Toledano QC conceded on behalf of the Buyer that in some circumstances a single person would count as more than one "DTH subscriber" for the purposes of the warranty. The concession, in its final form, was that the exception is limited to the case where one person pays for a package subscription at each of two separate households; and that such a person counts as two subscribers because (i) he would pay for two full price package subscriptions and (ii) he would also have a significant chance of paying for additional services at both locations, such as technical support. In opening submissions, the concession was not addressed to different households, but was that an individual who subscribes for two full package subscriptions (of which there were examples in the evidence) counted as two subscribers. This reflected the evidence of Mr Draca in analysing subscriber numbers on behalf of the Buyer, which treated such cases as involving more than one subscriber. This exception was also said to be justified on the basis that each subscription brought two full subscriber revenue streams and/or the opportunity for additional revenue from, for example, premium content.
47. In my view the concession of an exception, in either form, undermines the argument that a subscriber is a person. Once it is recognised that what are to be counted, in at least some cases, are subscriptions rather than persons, there is no basis for construing "subscribers" as a reference to persons. Moreover the argument that the exception can be justified because such a subscriber brings in two subscribers' worth of revenue conflates the Buyer's arguments that Second Television subscriptions do not count because they do not bring in enough revenue with the argument that a subscriber is a person, to which it has no logical application.
48. Moreover, if Second Television subscriptions are excluded, there might be excluded altogether a person who is a subscriber. Subject to the terms of the contracts under which promotions were made from time to time, a subscriber who had taken out a contract for a Second Television subscription was entitled to allow his primary subscription to lapse and use his Second Television subscription as his sole subscription. Although the subscriber needed an active subscription at the time he contracted for the Second Television subscription in order to qualify for the promotion, the two subscriptions thereafter operated completely independently because each operated from a unique smartcard. There was no mechanism by which the Second Television smartcard could be automatically deactivated when the first television smartcard was deactivated. Ms Vasiljevic and her team would have known this from their extensive due diligence and because this was the way the Company's operating systems worked. An example of this happening was a subscriber called Mr Dobrev, who for a period of 6 months was only paying for a Second Television subscription. On the Buyer's argument such a person would not count at all as a subscriber.
49. Once it is recognised that the warranty is concerned with the number of subscriptions, not the number of separate persons, the question becomes whether a Second Television subscription is a type of subscription which was intended to be counted separately under the warranty. I conclude that it was, for the following principal reasons.
50. In order to qualify as a subscription, the definition requires that there must be (i) a contract (directly or indirectly) with the Company (ii) to subscribe to the lowest package of the Company (iii) which is no more than 89 days in default. Second Television subscriptions fulfilled these criteria. They involved a separate contract with the Company, which involved a separate smartcard whose use was regulated independently of the primary subscription. The Company's conditional access system recognised each smartcard as a separate subscriber, no matter how many smartcards that individual had. The content might be different for each smartcard. A Second Television subscriber could add additional content to the Second Television subscription for an additional fee, just as he could to his first package subscription. One smartcard might be deactivated or cancelled without any effect on the other. The Company website shows that in September 2009 in order to qualify for the Second Television promotion then being advertised, it was necessary to enter into a two year contract. There was no obligation to enter into a two year contract on the first television package. Therefore it would have been obvious to someone reviewing the Second Television subscription on the website, as Ms Vasiljevic and Mr Draca say they did on behalf of the Buyer before entering into the 2009 SPA, that the Second Television subscription contract operated independently of the first television package.
51. There is room for debate whether it would have been correct to describe a Second Television subscription as a "package" in itself. There are some documents which were seen by, or available to, Ms Vasiljevic and those doing due diligence, which so described it. In other documents, notably the website and the document at Item 18.2 of Schedule 17 of the 2009 SPA, it is not included in what are described as the packages. I do not regard this debate as important, because the Second Television subscription did undoubtedly include a subscription to "the lowest package of the Company", i.e. the Economy Package: every Second Television subscription provided access to the channels which made up the content of this package. Irrespective of whether under a given Second Television promotion the subscriber had to choose a package independently of his primary package, or automatically obtained the same package as under his primary subscription (but in any event shorn of premium content), his Second Television subscription involved a payment for access to content which included that in the Economy Package. It was a contract for a subscription to the lowest package of the Company in the same way as any other package subscription. This is to be distinguished from add-on services, such as premium content, technical services or pay-per-view. It was not necessary to obtain a separate smartcard for these latter services, which were parasitic on the package which came with each smartcard, whether that for the primary package or that for the Second Television subscription.
52. The fact that a Second Television subscription was available at a price which was cheaper than the primary subscription does not detract from this analysis. The Company had a wide range of prices, discounts and promotions from time to time for different permutations of access to content, as is apparent from the 277 offerings listed in the subs_and_revenues_2009 spreadsheet.
53. Second Television subscriptions were treated as subscribers by the Company, and the Sellers, in their recording and calculation of subscriber numbers. The Company reported Second Television subscriptions as additional, separate subscribers to the television content providers and paid for content on that basis. This practice was known to the Buyer from the due diligence and spreadsheets exchanged as set out above. This was knowledge which in any event was reasonably available to the Buyer.
54. Accordingly, in the light of the background I have described, the expression "subscribers" is one which would have been understood by both the Sellers and the Buyer as including Second Television subscriptions.
55. This construction is supported by the fact that the Buyer knew, on the figures presented in the course of due diligence in 2009, and its own projections, that the Company did not claim to have as many as 80,000 subscriptions if Second Television subscriptions were ignored. The spreadsheets provided at the end of August 2009 identified about 75,000 subscriptions, maintained at about the same level in each month since February, if Second Television subscriptions were excluded. Although these figures give a picture at the end of August, not at completion 6 weeks later when the warranty was given, they reflect the expectations of the parties. I reject Ms Vasiljevic's suggestion that what she had been told about the HBO difficulties which had taken place earlier in the year led her to believe that the shortfall would be made up over that period. Ms Vasiljevic predicted in her September 2009 business plan that first subscription numbers would increase to 77,253 at the end of September 2009 and 78,407 by the end of October 2009. Whilst it is of course possible for a seller to give a warranty which both it and the buyer knows is incapable of being fulfilled, or is very unlikely to be fulfilled, a warranty is more naturally to be construed consistently with an intention and anticipation on the part of the parties that it will be capable, rather than incapable, of fulfilment.
56. It was contended on behalf of the Buyer that the commercial purpose of the DTH Subscriber Warranty was to identify the customer base of the Company which brought in both an average fee for the basic package and also the potential for additional revenue from add-on services. But the Second Television subscriptions equally had the potential for additional revenue from add-on services. There is no reason why the customer base should not be identified by including those as separate subscriptions. It was contended that it would flout business common sense if Second Television subscriptions could be counted as separate and additional DTH Subscribers because there was a substantial difference between the average amount of revenue generated from a package subscriber and the average amount of revenue generated from a Second Television subscription. But there was also a wide variation in the amount of revenue generated by the subscriptions to different packages, and the subscriber base comprised those whose contracts reflected 277 different offerings. According to the document in the data room, the Economy Package cost BGN 8.60 and the Special Practical Package BGN 19.80, with the HD Premium Package costing BGN 35.90. There was inevitably a potentially wide variation in the revenue being generated by each of the "DTH Subscribers", whether or not they included Second Television subscriptions. It was suggested that if Second Television subscriptions could be counted as separate and additional subscribers within the warranted 80,000, the Buyer would have no idea what proportion of the 80,000 were made up of package subscribers and what proportion were made up of Second Television subscriptions. That is true so far as the warranty goes, but it is equally true that the warranty provided no breakdown of the wide variety of terms which applied to package subscribers under historic offerings, or indeed how many of them were still active subscribers at all. The warranty of a single number to cover all subscribers, active or lapsed, was not itself a sufficient or useful tool for estimation of future customer and revenue development. That could only be based on a more detailed breakdown and analysis of the subscriptions and their revenue, which was the sort of analysis which was undertaken in the due diligence exercise and reflected in the August 2009 spreadsheets. No warranty was sought or given in relation to any of those details.

Hotel and cable subscriptions

57. A number of hotels had contracts for a package in order to provide the television content to their guests in each room. For reasons driven by the equipment used by the hotels, and perhaps by the hotel's desire to control access to channels amongst guests, the hotel typically bought one smartcard per channel (not per package) and the smartcard for each channel was inserted into a "head-end" system at the hotel, which enabled the signal to be re-transmitted to each room via cable lines. This was so whatever package was chosen by the hotel. In this way each hotel room had potential access to all the channels in the package bought by the hotel, which included at least those in the Economy Package. The hotel was invoiced monthly in advance on a per room basis. The number of rooms was the number of subscriptions recorded in the Company's accounting system. The number of rooms was reported as the number of subscribers to content providers, who were paid on this basis. Hotels therefore formed an exception to the general rule that the number of subscriptions was the same as the number of smartcards.
58. In respect of cable networks, the Company provided the cable operator with a separate smartcard for each ultimate customer, which was inserted into the customer's cable converter box in order to access digital content. The customer was treated like any other subscriber, save that his monthly subscription was paid to the cable operator and the latter paid the Company according to the number of smartcards provided.
59. The revenue from cable networks and hotels was in line with that brought in by individual subscribers. The Company invoiced the hotels and cable networks BGN 6,168.28 in August 2009 for 777 subscribers, giving an ARPU of BGN 7.94. This is marginally more than a Second Television subscriber and marginally less than an "Economy" subscriber.
60. All this would have been apparent to the Buyer from the due diligence exercises in 2008 and 2009.
61. Each hotel room which was paid for separately, and each cable customer whose subscription was paid for separately, therefore falls within the definition in the DTH Subscriber Warranty. Each is a separate subscription under a contract for access to content containing the lowest package, the Economy Package. The number of hotel rooms and the number of cable subscribers therefore count separately towards the 80,000 subscriptions required by the warranty.
62. The Buyer did not contend that a hotel, or a cable operator, constituted a single subscriber, or that, in the case of hotels, the number of subscribers or subscriptions should be taken to be the number of smartcards rather than guest rooms. The Buyer's case was that such subscriptions did not constitute subscriptions falling within the DTH Subscriber Warranty at all because (a) such payments were treated in the Sellers' Business Plan 2008 separately from "package" subscribers and (b) such payments did not generate the same basic amount as a package subscription and (c) such payments did not generate additional revenue from add-on services in the way that package subscriptions could. For reasons I have explained in relation to Second Television subscriptions, none of these is a reason which takes hotel or cable subscriptions outside the definition.

The data dispute

63. In order to understand the issues which arise, it is necessary to say something about the Company's computer systems, from which the data on both sides was drawn. The Company had a customer relationship management system ("CRM") which consisted of three interlinked systems: (a) the billing system, (b) the accounting system and (c) the uplink system.
64. The billing system was a bespoke software system used by the Company to record subscribers' contract and payment details, to generate invoices and to generate data which was fed to the uplink system to determine whether a smartcard was "active". The billing system did not contain any data on certain indirect subscribers, including hotels and those who subscribed through cable networks and other bulk distributors. The billing system therefore did not provide a complete picture of the Company's subscribers. The Company's records on these indirect subscribers were stored in the accounting system and the activation of their smartcards was not governed by the billing system.
65. The accounting system was used to produce the Company's management and audited statutory accounts. Each morning the billing system was reconciled against the accounting system. It was therefore the accounting system which contained the accurate and complete list of revenue from direct and indirect subscribers. The accounting system used commercially available and certified "Ajur" software. The Company's management of the accounting process was ISO certified. Therefore, unlike the billing system, the accounting system was complete and externally audited.
66. The final component of the CRM system was the uplink system. This uplink system controlled access to the Company's transmission broadcast signals to ensure that only authorised subscribers could access content. The input into the uplink system was made (i) electronically from the billing system for subscribers controlled through the billing system and (ii) manually for those indirect subscribers who were not within the billing system.

The data sources

67. Mr Sherbanov accessed the CRM system on 7 October 2009, the date of completion under the 2009 SPA, with the avowed purpose of identifying subscribers who qualified as fulfilling the DTH Subscriber Warranty. He downloaded four lists of subscribers recorded in the CRM system as active within 90 days, totalling 83,192 subscribers:

(1) A list of 79,772 active and deactivated subscribers (i.e. subscribers not in default of payment for 90 days or more) was taken from the Company's billing system as verified with the accounting system. This was referred to as the "Abonati" spreadsheet, after its title which is Bulgarian for subscriber. For each subscriber the Abonati spreadsheet includes the subscriber's name, his smartcard number, his personal ID number, and the number, date and amount of the most recent invoice paid. Second Television subscriptions were included as separate subscribers.

(2) A list of 777 subscribers with enabled smartcards in cable networks and hotels, was taken from the Company's accounting system and the uplink encryption system.

(3) A list of 2,020 subscribers with smartcards managed by Elco Star was taken from the Company's accounting system and uplink encryption system. Elco Star was one of a number of the Company's distributors. Under a contract dated 25 February 2008 it agreed to purchase 2,020 subscriptions lasting two years for a fixed lump sum subscription price, which was paid at the outset; Elco Star was then free to sell on the subscriptions for its own benefit and on its own promotional terms. Elco Star was thus obliged to pay 2,020 subscriptions, but the number of its customers who purchased, paid for and used the smartcards might vary from time to time and be a smaller number. Nevertheless the smartcards were all activated from the outset and were therefore treated as active on 7 October 2009. There were 12 duplicate smartcard numbers in Mr Sherbanov's list of Elco Star subscribers. Mr Sherbanov's evidence was that despite this, Elco Star paid for, and the Company activated, 2,020 smartcards for 2 years. Therefore, the Sellers contend, all 2,020 subscriptions should be counted. Since a reduction of 12 subscribers would not affect the result in this case, I shall treat there as being 2,020.

(4) A list of 623 VIP subscribers was taken from the Company's billing system as verified with the accounting system. At the trial it was common ground that VIP subscribers did not qualify for the purposes of the DTH Subscriber Warranty because they were not paying anything for their subscriptions. They were receiving the service for free rather than subscribing to it.

68. The Sellers' case is that Mr Sherbanov's lists of subscribers on the Abonati spreadsheet (79,772), of Elco Star subscribers (2,020) and of hotel and cable network subscribers (777), totalling 82,569 subscribers falling within the contractual definition, is the best evidence of the number of subscribers at the date of the DTH Subscriber Warranty.
69. The Buyer's data was procured by Mr Draca, SBB's software development manager. The Buyer was planning to advance a breach of warranty claim against the Sellers by the time it came to sell the Company to Newgrant, and for these purposes in November 2010 Mr Solak of SBB asked Mr Draca to make a copy of the Company's CRM System, which he sought to do. This data was downloaded to a DVD containing 366 tables of figures and information which Mr Draca said replicated that which existed on the system as at 7 October 2009 ("the DVD data"). Mr Draca explained that although the CRM System was in constant use between 7 October 2009 and November 2010 when he downloaded the backup, any new data added from time to time would not have changed the previous data which remained on the system, and in particular would not have altered the data relating to the position on 7 October 2009.
70. The DVD data was not, however, a replication of the CRM system, because it did not contain the programmes belonging to the system which linked the different parts of the data. It could not therefore be interrogated to identify the number of subscribers falling within the warranty without developing and applying programmes to interrogate the data. It was in this sense raw data.
71. Mr Draca developed and applied such programmes and interrogated the data. His approach was to start with all the bills in the system, of which he says there were 3,196,170. These came from the billing system. He then sought to apply filters in order to determine how many subscribers had a bill which had been paid in respect of any part of the period within 90 days prior to 7 October 2009 and included a subscription to a primary package (rather than to premium channels). He excluded Second Television subscriptions. In doing this exercise Mr Draca did not use the vast majority of the data on the DVD. His algorithms were only applied to interrogate 4 of the 366 tables. He applied the filters in the following stages:

(1) He first identified the unique subscription numbers in the system to identify the universe of subscribers, on the basis that each subscriber had a unique subscription number ("Filter 1"). Since each Second Television subscription involved an additional new subscriber number, this filter left in Second Television subscriptions. This yielded 123,486 subscriptions ever billed in the system.

(2) He then sought to identify bills in respect of any part of the period of 90 days prior to 7 October 2009 ("Filter 2"). This reduced the figure to 74,385.

(3) He then excluded bills for Second Television subscriptions ("Filter 3"). This reduced the figure to 65,917.

(4) Finally he excluded bills for premium channels ("Filter 4"). This reduced the figure by 523 to 65,394.

72. If, as I have found, Second Television subscriptions are to be included, Filter 3 falls to be disapplied, which results in a figure of 73,862. Moreover, because these are taken from the billing system, which is not where the subscriptions are to be found for hotel, cable network and the Elco Star subscriptions, these need to be added on. If, as I find, Mr Sherbanov's figures can be relied on to quantify these three additional categories, one reaches a total of qualifying subscriptions, on my findings and Mr Draca's analysis of the DVD data, of 76,396 subscribers. This is a shortfall of 3,604 from the warranted number of 80,000.
73. The Sellers asked Mr Kanchev to interrogate the DVD. Mr Kanchev is a programmer by training and worked as the Company's programmer as well as marketing manager. His evidence was that the DVD lacked a schema to show how the raw data should be rebuilt into a functioning database and also lacked the functionality to interrogate that rebuilt database. He tried to recreate the functionality by writing new programming code, but despite his familiarity with the system, he considered the resulting interrogation of the DVD to be inherently unreliable. He compared it to trying to fit together a broken jigsaw based upon educated guesswork. I did not find this to be an entirely fair criticism. Mr Draca disclosed the code he had written for the filters he had applied, and it was open to Mr Kanchev or others on behalf of the Sellers to identify any specific errors in the programming language, which was not an exercise they attempted. Instead they sought to undermine the results on the bases identified below.
74. The evidence and argument on the reliability of each side's figures ranged widely, and I will not attempt to address every last detail of it. I will instead discuss what I perceived to be the principal elements before stating my conclusions.

Discussion of the rival data sets

75. Mr Sherbanov's lists of subscribers were downloaded from the CRM system on 7 October 2009 for the express purpose of proving to the Buyer that the warranty had been satisfied. Mr Sherbanov explained the process by which he obtained the Abonati list of 79,772 subscribers from the billing system in some detail in evidence. He collated the previous 89 daily reports of active subscribers to give him a list of active subscribers and subscribers de-activated for less than 90 days. Mr Sherbanov was very familiar with the CRM system: he had had responsibility for the daily exercise of monitoring subscriber numbers by reconciling the data in the billing system, the accounting system and the uplink system. As to the other lists, Mr Sherbanov explained that he had obtained the list of cable network and hotel smartcards from the accounting system (where they were separately managed) and then established that these smartcards represented 777 different subscribers by cross-checking the payments in the accounting system against the VAT reports. The list of Elco Star smartcards was taken from the accounting system and uplink encryption system.
76. It was suggested on behalf of the Buyer that his evidence was unreliable because his lists included a few invoices dated the day after this exercise, 8 October 2009. I accept his explanation that these reflected a manual exercise of updating his Abonati list on 8 October 2009 in respect of a small number of indirect subscribers after he downloaded the data on 7 October 2009. They do not of themselves cast any doubt on the accuracy of his data.
77. I found Mr Sherbanov to be an honest, careful and impressive witness. That is not to rule out the possibility of human error in compiling the lists, or the possibility of human error in the loading of data on to the systems in the first place; evidence of such human error is to be found in the fact that Mr Sherbanov's lists contain 12 duplicate smartcard numbers for Elco Star subscribers. Nevertheless it was not suggested by the Buyer that Mr Sherbanov acted otherwise than in good faith in the exercise he performed, nor that his methodology was an inappropriate way of going about collecting the data. The Buyer did not suggest in cross examination any particular way in which Mr Sherbanov could have inadvertently over-counted subscriber numbers; the attack on the reliability of Mr Sherbanov's Abonati spreadsheet was at the general level that it was a manual exercise with the attendant risk of human error, by contrast with the electronic interrogation of the DVD which automatically produced the numbers relied upon by the Buyer. Although there is undoubtedly room for some human error on Mr Sherbanov's part in compiling the lists, and on the part of those inputting the data, the fact that there was no generic criticism of Mr Sherbanov's methods, or of his competence or honesty, lends considerable weight to the reliability of the data he produced given his familiarity with the system.
78. At a general level, the Sellers advanced three reasons why the DVD data should be regarded as unreliable. These were:

(1) The DVD lacked the software programmes to access and interrogate the data. As indicated above, I do not think this is a criticism of any substantial weight, given the disclosure by Mr Draca of the programme code he used to interrogate the data.

(2) The Buyer made two significant changes to the billing system in the early period of their ownership, which had the potential to affect the reliability of the data downloaded only in November 2010. First, between November 2009 and February 2010, the billing system was integrated with the SBB system in Serbia. According to the appendix to the Company's Annual Financial Report dated 31 December 2009, the Company then wrote off the billing system and software "after Management estimated that they were not going to bring economic benefits to the Company". This suggests the degree of integration with SBB's system was such that the billing system in its original form no longer existed. Secondly, the Buyer introduced to the Company a completely new approach to billing, changing it from payment in advance to payment after the event, which required changes to the billing system in March and April 2010. In my view these changes are capable of providing an explanation for the DVD data not being an accurate replication of the position at the date of completion of the 2009 SPA, as Mr Kanchev suggested. The Buyer submitted that the backup of the CRM data was a backup of the database to which the employees of the Company gave Mr Draca access in June 2010 when he was first investigating the number of subscribers at completion; and that accordingly the Company employees, and in particular Mr Georgiev, the Company's IT manager, evidently thought at that time that the system could properly be relied upon for this exercise; Mr Georgiev sent an email to Mr Draca explaining the algorithm that he should use on the system for the purposes of determining the number of subscribers at 7 October 2009. The Buyer submitted that if, as the Sellers contended, the Company's IT employees had altered the historic data such that the system could no longer be relied upon, Mr Georgiev would have said so rather than leading Mr Draca on a wild goose chase. This does not, in my view, quite meet the Sellers' point. Mr Georgiev would have supplied the algorithm requested for interrogation of the system, even if the changes meant that the results could not exactly replicate the position at completion. There is greater force in the Buyer's argument that two of the Sellers' own witnesses were Mr Georgiev (who did not give oral evidence) and a programmer at the Company during the relevant period, Mr Rujinov, neither of whom gave any evidence that there had been fundamental changes to the CRM System which would have rendered it unreliable for determining the number of subscribers as at 7 October 2009. Nevertheless I conclude that if there are other reasons to question the reliability of the DVD data, the changes in the system provide a plausible explanation as to how that may have come about.

(3) The DVD data is taken only from the billing system in the CRM, not the accounting system. This has two consequences. It does not include hotels and cable and Elco Star subscriptions, which were only recorded in the accounting system. Secondly it was suggested by the Sellers that because it has not been reconciled with the accounting system, it is less reliable than Mr Sherbanov's data, which had been so reconciled. But since the billing system was reconciled with the accounting system as a matter of routine, there is in my view nothing in this latter point which helps on the reliability of the rival sets of data.

79. In support of the reliability of the DVD data, the Buyer also made the general points that the DVD data contains a wealth of detail on every subscriber, by contrast with the much more limited information contained in Mr Sherbanov's lists; and that the Sellers did not attempt to engage on an analysis of the DVD data, despite their ability to do so with the benefit of the algorithms supplied by Mr Draca. There is force in both these points.
80. The Sellers submitted that their data is supported, and the DVD data undermined, by (i) the Company's VAT records; (ii) contemporaneous screenshots; (iii) the Buyer's contemporaneous conduct and documents; (iv) PwC's net debt investigations; and (v) the Company's audited revenue figures. I will address each in turn.

(i) VAT reports

81. The Company reported its revenue from subscribers to the Bulgarian tax authorities on a monthly basis. These VAT reports were the subject of detailed scrutiny by the independent auditor of the Company. Subscriptions were paid for in advance, following which the Company would issue an invoice. The invoices followed rather than preceded payment, so that they were in effect receipts. On the basis of this payment and invoice, the Company would submit VAT returns and pay VAT. The Company would not want to pay VAT on payments which it had not received and there was no benefit to the Company in inflating revenue when reporting, which would serve to increase the VAT payable. Therefore the VAT reports are potentially capable of providing strong independent and contemporaneous evidence of the Company's receipts from subscribers.
82. Of the 82,569 subscribers in Mr Sherbanov's lists, the Sellers were able to identify 81,383 (98.6%) of them in the VAT reports. The shortfall (1,186) was explained by the Sellers as being the result of their inability to obtain all the relevant VAT reports: they were able to obtain the VAT reports only for January 2008, July - December 2008 and for 2009. This provides some support for Mr Sherbanov's figures, but its weight is diminished by three factors: the VAT reports do not show, for payments made before 7 June 2009, whether the payment was in respect of a period falling within 90 days of completion; in relation to indirect subscribers paying under an invoice to their distributor, the VAT reports do not identify which indirect subscribers are covered by the payment; and the VAT reports do not identify whether a payment by one subscriber is in respect of more than one subscription. The VAT reports nevertheless provide some support for Mr Sherbanov's figures because, despite these qualifications, they are entirely consistent with Mr Sherbanov's lists.

(ii) contemporaneous screenshots;

83. There were two contemporaneous screenshots of the billing system taken prior to completion upon which the Sellers relied. Each contained totals of subscriber numbers under various headings.
84. One was an active subscriber screenshot as at 3 August 2009, which recorded that 75,783 subscribers in the billing system had made a payment for August 2009. To qualify as an active subscriber, it was necessary to have pre-paid for that month's service. On the first day of the month, the CRM system transmitted an authorisation message which activated the smartcard for the month. If the subscriber had not pre-paid, no authorisation message would be sent and the subscriber would be de-activated. All those 75,783 subscribers would therefore qualify for the warranty, having been active within 90 days of 7 October 2009. But this total would not count others who would additionally qualify as DTH Subscribers, namely:

(1) those indirect subscribers who did not feature in the billing system and therefore would not register on the screenshot, i.e. the 2,797 hotel, cable and Elco Star subscribers;

(2) subscribers who had been deactivated at the end of July/beginning of August 2009 but had not yet been re-activated for August 2009; the evidence suggested that many of the Company's subscribers only paid for their monthly subscription once they had been deactivated, and that the number of reactivations which occurred during the course of the month could be as many as 10,000; 3 August 2009 was a Monday, so there would be likely to be a number of subscribers, measured in thousands, who had yet to reactivate their smartcards for August 2009 at the time of the screenshot;

(3) subscribers who became deactivated between 7 July and 3 August 2009 and did not reactivate, who would qualify as DTH Subscribers through having been active subscribers during the 90 day period; and

(4) new subscribers taking out subscriptions between 3 August and 7 October 2009; the Buyer's business plan projected that the Company would gain roughly 2,000 subscribers between August and October 2009.

85. This screenshot is therefore consistent with and supports Mr Sherbanov's figures, and suggests a figure in excess of 80,000 for qualifying subscribers at 7 October 2009. By contrast, it cannot be reconciled with the Buyer's DVD data figure of 73,862 DTH Subscribers (including Second Television subscriptions), given that subscribers not in default of payment for 90 days or more as at 7 October 2009 must necessarily include those active on 3 August 2009. In fact, Mr Kanchev ran Mr Draca's algorithm to see what figure for active subscribers (including Second Television subscriptions) the DVD recorded for 3 August 2009 and found that the DVD only recorded 67,260 active subscribers on that date – an undercount of more than 8,500 against the screenshot. Mr Draca's explanation for the inconsistency between the DVD and the August 2009 screenshot was first that the screenshot might also include VIP subscribers, hotels and cable networks, and secondly that he did not know what algorithm had been used to obtain the figures. The first (if correct which was not entirely clear) could account for only about 1,400 subscribers. The second was not a point which undermined the reliability of the figures: they were taken from the billing system contemporaneously; the document on its face states what process of interrogation of the data was purportedly undertaken - it is a screenshot of "active subscribers by packages, promotions".
86. The second contemporaneous screenshot is dated 1 October 2009, shortly before completion, and records 63,764 "active subscribers" to the "Economical" package. It also records that there were 9,577 Second Television subscribers, although unlike the 3 August 2009 screenshot the words "included below" do not appear. This, together with the algorithm in the accompanying email from Mr Georgiev, suggests that in this screenshot the Second Television subscriptions were not included in the total of 63,764. This screenshot is therefore supportive of an active subscriber base of 73,341 as at 1 October 2009. Like the August screenshot, this would undercount the number of DTH Subscribers at 7 October 2009, because there would have to be added (i) subscribers who had deactivated since 7 July 2009 (ii) subscribers who in accordance with the common pattern had deactivated at the beginning of the month but reactivated by 7 October 2009 (iii) new subscriptions between 1 and 7 October 2009 and (iv) the Elco Star and hotel and cable subscriptions.
87. The two screenshots relied upon by the Sellers are consistent with and supportive of Mr Sherbanov's figures and inconsistent with the figures relied on by the Buyer drawn from the DVD data.

(iii) the Buyer's contemporaneous conduct and documents

88. The Company had a fully functioning CRM system on the date of sale to the Buyer. That system allowed the Buyer to see on any given day, at any given time, the exact number of subscribers. If, as the Buyer alleges, the Company had had some 15,000 subscribers fewer than warranted when it assumed control, the Buyer would have been able to see this from the CRM system immediately and without difficulty. The fact that subscriber numbers were readily available is also evident from the fact that the Company continued to report subscriber numbers to its content providers (as it was obliged to do) at the beginning of every month, an exercise which involved a daily calculation of subscriber numbers. Yet there is no contemporaneous record of any perception by the Buyer in the last quarter of 2009, or the first half of 2010, that there had been fewer subscribers than warranted; and no complaint of breach of the subscriber warranty was made to the Sellers until over a year later in December 2010.
89. The evidence of the Buyer's witnesses in seeking to explain this, particularly that of Mr Solak, Mr Draca and Ms Vasiljevic, was contradictory and unconvincing. I was not impressed by the suggestion that SBB was so distracted by numerous other problems with the Company in the post-completion period that it was prevented from analysing the subscriber numbers. The Buyer's suggestion that the shortfall was only discovered in June 2010 is improbable, if there really was the shortfall it alleges, because such a substantial shortfall would have been readily apparent from the CRM system without the need for a particularly detailed or sophisticated analysis of the figures.
90. Such evidence is in any event inconsistent with subsequent documents. By 9 July 2010, the Buyer was referring to its "plan" to rely on "reps and warranties claims" to avoid paying the Sellers the second tranche of the price; but there is no suggestion of an egregious breach of the DTH Subscriber Warranty in any internal documents at this stage (or indeed prior to this). The Buyer continued to investigate ways to avoid triggering payments to the Sellers in August 2010, but still there was no reference to the supposed discovery in June, or at any time, of a significant shortfall in subscriber numbers, even internally within SBB, let alone in correspondence with the Sellers. Even at this stage, the Buyer did not appear to consider subscriber numbers on completion as a potential breach of warranty. On 8 November 2010, Mr Knorr sent an email to Mr Solak and Ms Vasiljevic amongst others with a short-list of evidence to collect for the purpose of building a claim that the Sellers had breached warranties in the 2009 SPA. Not only was there no suggestion of the serious shortfall in subscriber numbers allegedly discovered over four months previously in June 2010, but Mr Knorr requested a search for evidence showing that the Sellers had been under-reporting the Company's 80,000 subscribers to the content providers; Mr Solak and Ms Vasiljevic did not respond that there was in fact a shortfall in the subscriber numbers warranted.
91. Ms Vasijevic's subs_and_revenues_2009_vv spreadsheet derived its figures from the Company's CRM system in August 2009 and suggested 84,114 active subscribers in July 2009, including 9,469 Second Television subscriptions. This is much more consistent with Mr Sherbanov's lists being correct than the DVD data.

(iv) PwC's net debt investigations

92. PwC conducted an extensive review of the Company's finances from October to December 2009 in order to compile the net debt report. Mr Tzvetkov considered that this had been done to "the usual standard that I would expect from an accounting firm of PwC's standing" and the Buyer "reviewed their analysis closely". Had there been a warranty shortfall of anything like 15,000 subscribers, I would have expected this to have been identified by PwC and the Buyer during the net debt review.

(v) the Company's audited revenue figures.

93. These were primarily relied on as evidence that there had been no decline of subscriber numbers in the months in 2009 leading up to the 2009 SPA, and they support the Sellers' contention that there was no such decline. Beyond that, I did not find them helpful in assessing the reliability of the rival data sets.

The individual subscriber "errors"

94. Each party also relied on an analysis of the position of particular named subscribers as examples said to prove, or at least suggest, an error in the other side's data. I found limited assistance in these examples of "errors", because there was always the possibility, indeed probability, of occasional mistake by human error in inputting so much data onto the system. However, to the extent they assist, they are in my view more supportive of the Sellers' case than that of the Buyer for the following reasons.

The "Dobrev error"

95. The Buyer asserts that 1,179 direct subscribers are included in the Sellers' list in respect of whom the DVD does not indicate payment in the relevant period. It was Mr Kanchev's evidence that 1,142 of these subscribers, including Mr Dobrev, are linked to invoices dated within 90 days of completion which have been cross checked against the Company's VAT reports. When the example of Mr Dobrev was put to Mr Draca in cross-examination, his response was to accept that there was an invoice in the CRM system for a payment by Mr Dobrev on 29 September 2009, but to contend that this was one of many "false invoices" created by the Sellers for BGN 35,000 in August 2009 and BGN 130,000 in September 2009. I have no hesitation in rejecting this allegation of deliberate manipulation of the CRM system, which was raised only for the first time in cross examination when Mr Draca was trying to meet the Sellers' case on the Dobrev error. The Dobrev error casts further doubt on the reliability of the DVD data.

The "Stoyanov error"

96. Mr Draca pointed out, correctly, that subscribers whose invoices were earlier than 90 days before completion would not necessarily qualify as DTH Subscribers, and gave as an example of one wrongly included in Mr Sherbanov's lists the case of Mr Stoyanov. Mr Sherbanov's list counted both Mr Stoyanov's smartcards as active and qualifying as two subscriptions. The DVD data suggested that Mr Stoyanov was in default on his second smartcard from 28 April 2009. The invoice said to support Mr Sherbanov's inclusion of Mr Stoyanov's two active smartcards was a single invoice of 1 February for a single 12 month subscription. This therefore appeared at first sight to support the Buyer's case of an error in the Abonati spreadsheet. The Sellers wrote to the distributor who handled Mr Stoyanov's subscription seeking clarification, and the distributor responded in a letter with information which, together with Mr Kanchev's subsequent evidence, provided an explanation which was to the effect that the 12 month subscription had been allocated across the two smartcards so as to equate to 5 months from 1 February 2009 on Mr Stoyanov's first smartcard and 7 months from 28 April 2009 on the other. The Buyer complained, with some justification, that this evidence emerged very late and in a form which could not fairly be tested. It is, however, a possible explanation for the so called "Stoyanov error", and renders the "Stoyanov error" a factor of little weight in the debate as to the reliability of the rival data sets.

The "Atanasov error"

97. Mr Atanasov is an example of an indirect subscriber, whose individual subscription is not identified by name on the distributor's invoice, which covers multiple subscriptions. According to the DVD data, Mr Atanasov made payments as a direct subscriber only up to 27 February 2009, and was thereafter in default on his "Basic" package (until 27 June 2010). By contrast, Mr Sherbanov's list includes Mr Atanasov against a group invoice for BGN 1,999.61 dated 5 February 2009 which covers a number of unidentified indirect subscribers. Mr Kanchev identified Mr Atanasov's subscription as included within the invoice. Mr Draca, on the other hand, said that the invoice did not include Mr Antonasov's subscription but related to 86 other subscribers. In this he was relying on an algorithm he said he had applied to the DVD data and the suggestion arose in cross examination so that it could not be fully tested (and was not put to Mr Kanchev in cross examination). It is impossible to judge who is right on that question. If Mr Kanchev is right, the issue is simply whether Mr Atanasov's subscription was for a period which brought him within the 90 days. Given the size of the group invoice, it is entirely feasible that it did. This so called "error" does nothing to undermine the Sellers' case (or indeed that of the Buyer).

The "Lozanov error"

98. Mr Draca identified 2,279 indirect subscribers in the Sellers' list linked to group invoices dated within 90 days of completion, who he says should not be counted as DTH Subscribers on the basis that (a) they could have made payment to the distributor much earlier than the distributor paid the Company and/or (b) Mr Sherbanov's list wrongly links the subscriber to the invoice. Mr Draca cited Mr Lozanov as the single example of this supposed error.
99. As to the first criticism, Mr Sherbanov's unchallenged evidence, which I accept, is that distributors' invoices for payments made to them by the indirect subscribers came within the third or fourth day after receipt of the subscriber's payment. Invoices within the 90 days are therefore unlikely to include non qualifying subscribers. As to the second, similar points arise as for the so called "Atanasov error", and no safe conclusions can be drawn.

The "refund" error

100. The Buyer suggested that there were about 100 invoices for negative numbers for subscriptions recorded in the Abonati spreadsheet, and that this was an error because subscribers who had been offered refunds should not count as "DTH Subscribers". This was raised for the first time in cross examination of Mr Sherbanov, who was unable to explain the negative entries. The Sellers' explanation in final submissions was that the Abonati spreadsheet records the latest invoice in respect of each subscriber, and the fact that this might be a credit note is in itself neutral; a subscriber might be active or have been deactivated at the time they received a credit note. So whilst it is true that the refund cannot itself prove that a subscriber is active (any more than an invoice for payment from before the 90 day period), it does not follow that the subscriber was not active. This explanation was supported by particular subscriber examples and seems to me to meet the Buyer's point. The point was not pressed by the Buyer in final speeches because its own lists included people with negative invoices.

Conclusion on the rival data sets

101. I cannot conclude with any degree of confidence that either data set is entirely reliable and accurate. Nevertheless, I regard the subscriber numbers contained in Mr Sherbanov's lists as likely to be distinctly closer to the true figure than those based on the DVD data, for the reasons discussed above. The burden of proof rests upon the Buyer to prove the breach of warranty alleged. It is for the Buyer to prove that the qualifying number of DTH Subscribers was fewer than 80,000. On my findings as to which subscriptions qualify, Mr Sherbanov's lists give a total of 82,569 qualifying subscribers. The Buyer has failed to satisfy me that the DVD figures are more reliable than the Sherbanov lists to the extent of a difference of some 2,500 subscribers, which it would be necessary for the Buyer to establish in order to prove a breach of the warranty; on the contrary, I find on the balance of probabilities that there were at least 80,000 qualifying DTH Subscribers at 7 October 2009 and that there was no breach of the DTH Subscriber Warranty.

Other warranty claims

102. Paragraph 1 of Schedule 2 of the 2009 SPA provided that:

"The information disclosed in the Due Diligence Documents is in all material respects true and correct, and such information fairly presents the legal and financial situation of the Company. All material facts about, or circumstances relating to, the assets, business or financial condition of the Company have been fairly disclosed in the Due Diligence Documents.

103. The Buyer alleged that in breach of this warranty, the Sellers failed to disclose that the number of subscribers had been in steady decline since at least early 2009; and had, on the contrary, presented the subscriber base as relatively stable notwithstanding the difficult economic climate. This was not alleged to give rise to any loss independently of breach of the DTH Subscriber Warranty, and so I can deal with it briefly. My conclusion on the evidence was that there was no such decline.
104. In addition, the Buyer also alleged breaches of warranty in respect of undisclosed litigation, customs irregularities and broken set top boxes. The Sellers disputed that there were any such breaches. The Buyer did not allege, or seek to prove, that it had suffered any losses from these alleged breaches. It made clear that it asserted these breaches as relevant only to the quantification of loss for the alleged breach of the DTH Subscriber Warranty, by way of supporting its case that the general state of the Company was such that it could not be turned around with the lower revenue streams produced from a subscriber base of only 65,394 and, therefore, that the shares were worthless. In the light of my conclusions on the DTH Subscriber Warranty issue, I do not need to address these allegations.

Conclusion on warranty claims

105. It follows that the Buyers have failed to establish any breaches of warranty giving rise to a loss; and that the Sellers are entitled to the tranche 2 balance of the purchase price, in the sum of €4,194,000.

The Net Debt Claim

An outline

106. Pursuant to clause 7.1 and para 9(e) of Schedule 2 of the 2009 SPA, the Sellers had to provide a warranty as to the net debt of the Company as at 31 August 2009 and at completion. Schedule 3 of the 2009 SPA contained a schedule of figures under ten headings which gave a total warranted net debt of BGN 42,875,990 (the "Sellers' Net Debt Amount").
107. Clauses 3.4 to 3.6 of the 2009 SPA provided for a post-completion adjustment of the net debt, with a dispute resolution mechanism in the event of a disagreement. There were to be four stages:

(1) First, PwC were to carry out a review of the net debt of the Company "as an expert" and deliver the results of the review to the Buyer by 1 December 2009: clause 3.4.

(2) Second, the Buyer was promptly to notify the Sellers of the "amount resulting from PwC's review" (the "Buyer's Net Debt Amount"): clause 3.5.

(3) Third, if the difference between the Buyer's Net Debt Amount and the Sellers' Net Debt Amount was greater than €100,000, the parties had to endeavour to agree the Net Debt Amount within 2 days but in any event not later than 14 December 2009: clause 3.6(2)(a).

(4) Fourth, if the parties failed to agree, the parties were promptly and jointly to nominate one of the "Big Four" accounting firms to calculate the Net Debt Amount based on an audit of the net debt at 31 August 2009; the accounting firm's audit was to be undertaken at the Sellers' expense, and its decision was to be final and binding: clause 3.6.2(b), (d).

108. There were two clauses of the 2009 SPA dealing with the provision of information which are relevant to the net debt adjustment process.

(1) Clause 3.4.2 provided:

"In so far as they are able, the parties shall each provide, and procure that the Company provides, PwC with all information relating to the Company which PwC reasonably requires, including but not limited to all relevant breakdowns and analytical information related to the Net Debt."

(2) Clause 5.3 provided:

"The Buyer undertakes to provide full access to the Sellers to all accounting information and documents on the grounds of which the Net Debt review and/or audit pursuant to Clauses 3.4, 3.5 and 3.6 is carried out."

109. PwC started the net debt review shortly after completion, and provided the Buyer with a draft report running to 42 pages dated 16 November 2009. That report included adjustments which it was thought the Sellers themselves were likely to accept (labelled collectively "conservative" adjustments) and other "possible" adjustments (labelled collectively "aggressive" adjustments). It also included certain possible tax liabilities which the authorities might claim if detected.
110. Following discussions between Mr Babace of Mid Europa and Ms Nalbantova of PwC, PwC produced a single page net debt statement ("the PwC Net Debt Statement"), which listed the 22 items in respect of which PwC proposed an adjustment from the Sellers' Net Debt Amount, with a very brief explanation for each. For the most part it adopted the figures described in the draft report as "aggressive". It identified the adjusted net debt as BGN 50,898,000, an upwards adjustment of BGN 8,023,000 (some €4m) from the Sellers' Net Debt Amount. In accordance with the 2009 SPA, the Buyer informed the Sellers of the amount resulting from the PwC review by a letter dated 1 December 2009, which therefore became the Buyer's Net Debt Amount. This letter enclosed the PwC Net Debt Statement, and additional supporting enclosures which gave further information on most of the 22 items.
111. The Sellers rejected the Buyer's Net Debt Amount in its entirety. There followed a series of technical meetings with PwC to discuss the PwC Net Debt Statement and negotiation meetings between the Buyer and the Sellers to seek to agree the Net Debt Amount.
112. The Sellers' representatives in all of the technical meetings with PwC included Ms Dimitrova and Ms Angelakieva, the Company's Chief Accountant between 2002 and 2009. Four financial and tax advisors of the Sellers attended some of the meetings (Ms Tomova, Mr Barutev, Mr Pavlov, and Ms Bakalova). There are contemporaneous lawyer's minutes for the majority of these meetings, and the negotiation meetings, prepared by the Buyer's lawyers, DGKV. The content of these minutes was, in part, disputed by the Sellers. They were prepared by DGKV not just for their own clients, the Buyer, but with a view to circulation to both the Buyer and the Sellers at the time, and they were circulated contemporaneously to both. This makes them more likely to be materially complete and accurate. The minutes show (and this much was common ground) that, prior to the Net Debt Agreement, representatives of the Sellers made repeated complaints about their access to information and repeatedly asserted that the PwC Net Debt Statement was wrong. The Sellers relied upon their contractual right to information, including that contained in clause 5.3 of the 2009 SPA.
113. The negotiations were conducted by Ms Ivanova and Ms Dimitrova on behalf of the Sellers, and Mr Tzvetkov on behalf of the Buyer, at meetings on 11 and 14 December 2009 and, following further technical meetings with PwC held at the Company's offices on 15 and 16 December 2009, by telephone. In the final stages, the Sellers offered an adjustment of BGN 2.5m but the Buyer wanted an adjustment of BGN 3m. In the end, they agreed to halve the difference and settled on an adjustment of BGN 2.75m. The parties then entered into the Net Debt Agreement, dated 28 December 2009, by which the parties agreed that the net debt of the Company was BGN 45,625,990 (i.e. BGN 2.75m more than the Sellers' Net Debt Amount as warranted in the SPA, but a substantially smaller adjustment than the BGN 8m adjustment originally proposed in the PwC Net Debt Statement on 1 December 2009); and made the consequent adjustment to the tranche 2 purchase price (lowering it by €1,406,000, which is equivalent to BGN 2.75m). The figures demonstrate that the final adjustment involved some give and take, and represented a compromise, freely entered into by both sides.
114. The Sellers seek to avoid the effect of the Net Debt Agreement by advancing three bases of claim, namely misrepresentation, breach of clause 5.3 of the 2009 SPA, and breach of clause 3.4.2 of the 2009 SPA. Before addressing them I must set out the course of the discussions in a little more detail.

Narrative

115. Following notification of the Buyer's Net Debt Amount on 1 December 2009, the exchanges started with a fax dated 4 December 2009 from the Sellers rejecting it in its entirety. There was then a technical meeting held between PwC and the Sellers on 9 December 2009. It is clear from the contemporaneous documentation that the Sellers continued to challenge the PwC Net Debt Statement at that meeting. Ms Ivanova asked for an explanation of each of the individual items, but was not satisfied with the answers she was given. The day after the meeting, the Buyer sent a fax to the Sellers requesting that they provide a list of detailed questions in relation to the Net Debt Statement, in order for there to be a further working session with PwC in advance of a proposed meeting between the Buyer and the Sellers on 11 December 2009. Ms Ivanova's evidence was that it was agreed at the 9 December meeting that the Sellers' representatives could visit the offices of the Company at 10am the next day, but that when the Sellers' representatives (not including Ms Ivanova) visited the premises of the Company at 10am on 10 December 2009, they were refused access. Mr Tzvetkov explained that he was unaware of this alleged incident.
116. Ms Ivanova sent an email shortly before noon on 10 December stating that:

"…in accordance with article 5.3 of the SPA dated 7 October 2009 we require immediate execution of our right of full access to all accounting information and documents on the grounds of which the Net Debt review and/or audit pursuant to Clauses 3.4, 3.5 and 3.6 was carried out. Our representatives will be at [the Company's] premises today at 2pm. Bulgarian time."

Mr Tzvetkov replied immediately stating that he thought PwC had indicated that they were not available that day and this was rather short notice. He proposed a meeting the next morning.

117. Ms Dimitrova then sent Mr Tzvetkov a list of documentation to which she wanted access. This is in essence the documentation to which the Sellers now complain that they did not have access, including, in particular, the entire CRM system. Mr Tzvetkov forwarded this email to Mr Babace who in turn requested that Ms Nalbantova of PwC prepare the requested information for the meeting the next morning with the Sellers. Ms Nalbantova replied, itemising the documents which PwC would provide the next day to the Sellers, being that information relied on by PwC for their net debt review. Ms Nalbantova queried whether the Sellers were entitled to copies of the documents or were only to be allowed access to the information during the proposed technical meeting. Mr Babace replied, explaining that hard copies were fine, but the Buyer would prefer not to provide the Sellers with electronic copies of large data files where they would inadvertently obtain more information than the Sellers were entitled to as part of this exercise; in such a case, printouts of relevant items or strictly supervised access would be preferable.
118. Ms Nalbantova also emailed to the Sellers a release letter which explained that Mid Europa and SBB had confirmed that PwC could provide access to, or a copy of, their own analyses used for and/or included in PwC's draft report dated 16 November 2009; and PwC would do so if the Sellers agreed to hold PwC harmless against liabilities for providing such information.
119. Mr Tzvetkov emailed Ms Ivanova and Ms Dimitrova to explain that PwC was working to provide the data requested, and that there could be a working session at 10am the following day at PwC's offices. Ms Ivanova replied stating that there had been a misunderstanding and that the Sellers wanted full access to the "accounting system, billing system etc." The Sellers once again expressly relied on clause 5.3 of the SPA. Mr Tzvetkov replied at 8am the following day, explaining that clause 5.3 gave the Sellers a right of access only to the accounting information and documents based on which PwC performed the net debt review, which PwC would be providing at their offices that morning.
120. In the end, the Sellers did not attend the meeting with PwC scheduled for 10am on 11 December 2009. The Sellers did, however, attend the meeting with the Buyer scheduled for noon that day at the Second Claimant's offices in Sofia. Mr Tzvetkov attended by telephone. There is a full minute of that meeting drafted by DGKV, which I treat as materially accurate. The Sellers' legal representative, Mr Tabakov, complained that no one from the Buyer had attended in person. Ms Dimitrova complained that:

"…the Buyer had denied her access to the information and documents underlying the Net Debt calculation of PwC as she was not allowed to enter the building of the Company."

Mr Tzvetkov explained that the Buyer had facilitated access to the information required in compliance with the 2009 SPA. He explained that the Buyer had arranged a meeting with PwC at 10am that day to provide the Sellers with

"1) the accounting information originating from the Company based on which the Net Debt review was carried on as specified in Section 5.3 of the Share Sale Agreement; and 2) the procedure for release of additional PwC analysis and calculations relating to the Net Debt".

The Sellers' lawyer explained that the Sellers did not want access to the PwC analysis on the Net Debt, but just wanted the underlying accounting information and documents originating from the Company. Mr Tzvetkov explained that this information had been made available by PwC that morning and confirmed that the information could be provided again at a time when the Sellers could make themselves available. A meeting with PwC was arranged for 2.30pm on Monday, 14 December 2009 and a meeting with Mr Tzvetkov was also arranged for that day. Mr Tzvetkov agreed that he would come to Sofia to attend the meeting in person.

121. In the event, a meeting was held at the offices of PwC at 3.15 pm later on the same day, 11 December. DGKV's minutes of the meeting, which I treat as materially complete and accurate, record the following. Ms Dimitrova attended with Ms Angelakieva, Ms Tomova and Mr Barutev; Mr Babace of the Buyer attended by telephone. Ms Nalbantova of PwC made available all the documents containing the grounds on which PwC had conducted their net debt review and explained to the Sellers' representatives that they could review all the documents and take notes. Certain documents (which are itemised in the minute) were reviewed by the Sellers' representatives. The Sellers' representatives took notes and, at the end of the two and a half hour meeting, asked to be allowed to review the rest of the documents at 10am on Monday, 14 December 2009, which was agreed.
122. Ms Ivanova's evidence was that the Claimants were provided with "virtually no underlying documents" at this meeting. Ms Ivanova did not attend the meeting but obtained her account from Ms Dimitrova. I prefer the evidence of DGKV's detailed contemporaneous minutes of the meeting which record that the Sellers were provided with a large amount of documentation concerning PwC's adjustments to the Net Debt. This is consistent with and supported by the email PwC had sent to Mid Europa the previous day, setting out a long list of the information which it was intended to provide to the Sellers at that meeting; there is no reason why PwC should not have done what they said the day before they intended to do, and no such reason was suggested on behalf of the Sellers. Ms Dimitrova confirmed in cross examination that they had been shown detailed figures on screen, but said that her real complaint was that they had not had access to the systems and could not put the figures provided into Excel files in order to conduct their own analyses of them.
123. A further short meeting was held at the offices of PwC at 10.15am on Monday, 14 December 2009. The minute of that meeting records the following. Ms Dimitrova began the meeting by asking what criteria and formulae had been applied to the information in the billing system. Ms Nalbantova explained that this was a request for analysis, rather than a request for information (the Claimants had refused to sign the PwC release letter pursuant to which PwC would provide their analyses to the Claimants and the Claimants would hold PwC harmless from liability for these analyses). Ms Dimitrova and Ms Angelakieva reviewed various underlying documents and took notes (Ms Ivanova's second hand evidence was that they were not allowed to take notes, but I prefer the contemporaneous minutes of the meeting as an accurate record). Ms Dimitrova and Ms Angelakieva stated that there were discrepancies with the documents but did not specify further details.

The 14 December meeting

124. At 2.30pm on the same day, 14 December 2009, the parties held a meeting in order to seek to agree the net debt adjustment. It was attended by Ms Ivanova and Ms Dimitrova, together with their financial consultants, on behalf of the Sellers, and by Mr Tzvetkov on behalf of the Buyer. Each side had lawyers attending. Ms Nalbantova and an assistant from PwC were present. There are contemporaneous minutes of that meeting prepared by DGKV. During the course of the meeting, there was a break-out meeting in another room attended only by the principals (the "Break Out Meeting"), of which there is no contemporaneous record.
125. The minutes of this meeting record that Ms Dimitrova started the discussion by stating that the Sellers completely disagreed with the Buyer's Net Debt Amount. She also complained of the lack of direct access to information. The scope of Ms Dimitrova's complaint was set out in the minutes in the following terms:

"Mrs Dimitrova said a month ago she had asked for full access to accounting information and such access was refused to her. She said Mrs. Dragica Pilipovic called her in that respect saying she had no right to visit the Company. She further claimed the information the Sellers were provided with at the meetings held on 11 December 2009 at 2 p.m. Bulgarian time and on 14 December at 10 a.m. Bulgarian time at the offices of PwC was only limited information originating from the Company. She insisted in addition the Sellers to be provided with access to the method that PwC used in the Net Debt review".

It can be seen that there were two parts to this complaint. First, Ms Dimitrova was complaining that she had not been provided with access to the Company; secondly, Ms Dimitrova was complaining that she had not been provided with PwC's analysis. The Buyer's case is that the Sellers were not entitled to either under the terms of clause 5.3 of the 2009 SPA. Mr Tzvetkov explained that the Buyer was willing to give the Sellers more time to review the information at the Company's offices and was content for the period for this review to be extended beyond the contractual limits up until 17 December 2009. Ms Dimitrova again complained about not being given access to the method PwC had used for its review but her financial consultant stated that the Sellers did not need this information and the source material would be sufficient. Ms Dimitrova repeated at several stages that her opinion was that either the information provided to PwC was not sufficient, or the result of its processing was incorrect. Ms Ivanova stated that the Sellers would look for an arbiter because they did not agree with PwC's calculation. Mr Tzvetkov explained that: "[the Buyer] trusts PwC and the team of Mrs. Nalbantova without any reserves and that they professionally did their job. It was the right of the Sellers to disagree with the results." Mr Tzvetkov explained that the Buyer had stated its proposal clearly and expected the Sellers to state whether or not they agreed with it. The Sellers' legal representatives asked the Sellers whether they found any grounds to negotiate and said that, if the Sellers totally disagreed, the parties should proceed to the arbitrating accountant procedure. Ms Dimitrova then asked Mr Tzvetkov to go out of the room for five minutes in order to discuss the Buyer's proposal and the Net Debt Amount. Ms Ivanova, Ms Dimitrova and Mr Tzvetkov did so and had a discussion in a separate room at the Break Out Meeting.

126. There is a substantial dispute over what was said at the Break Out Meeting between Ms Dimitrova, Ms Ivanova and Mr Tzvetkov. In particular, the evidence of Ms Dimitrova and Ms Ivanova was that Mr Tzvetkov immediately stated that all of the 22 adjustments save for items 1, 2, 4 and 5 were incorrect, and could be removed entirely from the discussion; and that Mr Tzvetkov then insisted that item 1 was not up for discussion, saying that "there was no doubt that figure was correct", and was prepared to discuss only items 2, 4 and 5. Mr Tzvetkov's evidence, by contrast, was that he set out his bottom line which did not include item 1, but included a number of other items. The largest of these were items 2, 4 and 5, but he did not concede all other items. His evidence was that he explained that, unless the Sellers could agree on items 2, 4 and 5, there was little point negotiating and the parties should proceed to instruct an arbitrating accounting firm. His evidence was that he did not say that any items were correct because he was not in a position to do so: these items were the result of many layers of work and he had not performed the review himself, nor had access to the primary accounting materials.
127. The only aspect of the disputed evidence as to what happened at the Break Out Meeting which I need to resolve is what was said about item 1. On this issue I prefer the evidence of Mr Tzvetkov, because it is supported by contemporaneous documents showing his negotiating position going in to the meeting, including in particular a "0" against item 1 suggesting that he would not insist on any value being attached to this item, and by the fact that PwC had already indicated in their draft Report that this was not one of the stronger items. It is also supported by the fact that the minute records Ms Dimitrova as saying when they returned to the full meeting that the items on which they had failed to agree were items 2, 4 and 5; had Mr Tzvetkov insisted that item 1 was correct and non negotiable, that would not have been agreed by Ms Ivanova and Ms Dimitrova, and would have been included amongst the items of disagreement identified when the full meeting resumed.
128. Following the Break Out Meeting, the minute records that:

"When they returned Mrs. Dimitrova said they made an agreement. She said that the points on which they failed to agree were items 2, 4 and 5 in the Buyer's Net Debt Amount attached to the Buyer's letter dated 1 December 2009. She said they agreed on all other elements of the Net Debt. Mrs Dimitrova said the Sellers need additional time to review items 2, 4 and 5 of the Buyer's Net Debt Amount.

Mr Tzvetkov explained it was agreed the Sellers to receive extended access to the information necessary to assess items 2, 4 and 5 of the Net Debt calculation in the ITV offices under the supervision of DGKV and PwC. He said the Sellers might take notes but might not take documents out of ITV offices. He said the Parties agreed a conference call on Thursday when would be the deadline for the Sellers to express their position on the Buyer's Net Debt Amount and whether the Parties may reach an agreement or shall proceed to the arbitrating procedure. He said that it was agreed meanwhile the Parties to start working on the scope of work of the Arbitrating Accounting Firm and to approach the possible Arbitrating Accounting Firms (one of the "big four")."

129. Ms Dimitrova again asked whether they would receive information about the "method PwC used in the Net Debt review". Ms Nalbantova, of PwC, stated that they had explained the method and, if the Sellers required further information, they should sign the release letter. The Sellers' legal representative agreed that the release letter was a standard risk management procedure for the big four accounting firms but explained that she had advised the Sellers not to sign a release letter and, instead, "to use the source information for their calculations."
130. Following the meeting, Mr Tzvetkov sent an email to Mr Nikolai Nikolov (the Company's current Chief Accountant), copying in both Ms Nalbantova of PwC and Ms Ugrinova of DGKV, which stated:

"Pls be advised that following the meeting with Elena Illieva Ivanova (principal of CIG) and Tzvetelina Dimitrova (principal of Bikam),…… feel free to provide them and their advisors access to primary financial information as at 31 August 2009 which was used for the post closing Net Debt review (including system access). Such access should be given until end of the day Wednesday (Dec-16) and is to take place on the premises of ITV at a time of your convenience and in the presence of a representative of DGKV."

131. Mr Nikolov replied asking whether, if the Sellers asked for clarifications of the information, he should provide these clarifications or just allow them access to the systems; and whether the Sellers were allowed to download, copy and print the information. Mr Tzvetkov forwarded this email to Ms Nalbantova of PwC saying that he would answer positively but asking whether she had a preference for the opposite. Mr Tzvetkov was not able to find a reply to this email but, about two hours later, he emailed Ms Ivanova explaining that he had arranged extended data access at the Company until 1600 on 16 December 2009. The Sellers adduced evidence in the form of a witness statement from Mr Nikolov who stated that he does not think that Mr Tzvetkov responded to his two questions and so he was extremely cautious at the subsequent meetings on 15 and 16 December 2009.
132. The Sellers' representatives attended the offices of the Company on the morning of 15 December and again in the afternoon, and on the following morning, 16 December 2009. The meetings were attended by Ms Dimitrova and Ms Angelakieva on behalf of the Sellers and Ms Nalbantova of PwC, and at times, Mr Nikolov, Mr Rujinov and Mr Georgiev, employees of the Company. During the meetings the Sellers were given some access to the underlying accounting information, but contend that the limited material available made it impossible for them to verify PwC's adjustment. At one stage when the Sellers' representatives were seeking confirmation of the accuracy of the figures in relation to deferred revenue (item 4), Ms Nalbantova invited into the meeting Mr Georgiev and Mr Rujinov, who confirmed that the data relied on was accurate and had been properly extracted from the billing system.
133. After the meetings, Ms Ivanova and Ms Dimitrova contacted Mr Tzvetkov by telephone to negotiate the net debt adjustment. The Sellers' starting point was BGN 2.5m and Mr Tzvetkov's was BGN 3m. In the end, the parties decided to compromise on the figure of BGN 2.75m. This agreement was recorded in an email exchange on 19 and 20 December 2009, after which the parties negotiated the terms of the Net Debt Agreement. The Net Debt Agreement was executed on 28 December 2009, stating amongst the recitals that the Buyer's Net Debt Amount had been:

"…determined pursuant to the procedure set forth in Section 3.4 of the Share Sale Agreement and was communicated by the Buyer to the Sellers by a letter dated 1 December 2009."

134. Both Ms Ivanova and Ms Dimitrova stated that, between the technical meetings with PwC on 15 and 16 December 2009, and the parties agreeing to the Net Debt Agreement, they were still unhappy about their access to information. Ms Ivanova explained that her accounting team told her that they had not seen all the information. Ms Dimitrova stated that she had not been able to run all of the analysis that she wanted to run, and so was left with concerns and was disturbed about the situation. Ms Angelakieva explained that, in her opinion, the Sellers had been provided with a small number of documents and reviews which had been carried out in a format which they could not cross check, and that they had lacked sufficiently clear answers about the methodology used. Nevertheless, the Sellers made a conscious choice, for commercial reasons, to negotiate and agree the net debt with the Buyer, instead of either proceeding to instruct an arbitrating accounting firm or making any further complaints about their lack of access to information.

Misrepresentation

135. The Sellers pleaded five misrepresentations.
136. The first was that Mr Tzvetkov represented at the Break Out Meeting on 14 December 2009 that item 1 of PwC's adjustment was correct. This misrepresentation plea fails for each of the following reasons:

(1) No such representation was made by Mr Tzvetkov: see above.

(2) Had it been made, it would not have been an actionable representation of fact. The parties knew that the background to the meeting was that PwC had determined the net debt adjustment in accordance with their Net Debt Statement of 1 December 2009; they were engaged in a commercial negotiation in seeking to agree an adjustment, knowing that in the absence of agreement the matter would have to be referred for determination by one of the "big four" accounting firms. Ms Ivanova accepted in evidence that she was well aware that Mr Tzvetkov had not performed the PwC net debt review himself, and had not attended any of the technical meetings concerning that review. The Sellers were fully aware that Mr Tzvetkov's role was to negotiate on behalf of the Buyer and that he was not in a position to verify the accuracy of PwC's accounting exercise. In these circumstances, had there been a statement by him that an item was "correct", it would have been no more than a contention or argument and would have been reasonably understood by the Sellers' representatives as such, given the context in which it was made. Such a contention or argument does not amount to an actionable representation: see Kyle Bay Ltd v Underwriters Subscribing under Policy No. 01957/08/01 [2007] Lloyd's Rep IR 460,466, at [30]–[32] per Neuberger LJ.

(3) Had the alleged representation been made and been actionable, the Sellers would not have relied upon it. Ms Ivanova and Ms Dimitrova always believed that item 1 was not correct and that the adjustment suggested by item 1 was not justified. It was the Sellers' case that they had entered into the Net Debt Agreement believing the alleged representations that, at least, items 1, 2, 4 and 5 of the PwC Net Debt Statement were correct. However the evidence of the Sellers' witnesses at trial did not support this case. The Sellers would not have believed any such representations even if they had been made. They were doing everything that they could to demonstrate that the items were not correct. They decided to compromise not because they placed any reliance on what Mr Tzvetkov may have said about the correctness of the figures, which they continued to believe were incorrect, but because they thought that they were unable to demonstrate the incorrectness of the figures by reference to documentation which would convince the Buyer, PwC or an arbitrating accounting firm.

137. The second misrepresentation alleged by the Sellers is that at, and after, the meeting on 14 December 2009, Mr Tzvetkov represented that items 2, 4 and 5 of PwC's adjustment were correct. This misrepresentation plea fails for each of the following reasons:

(1) There was no such representation. Mr Tzvetkov did not concede that any of the items were to be discounted but suggested that unless agreement could be reached on items 2, 4 and 5, no agreement would be reached on the adjustment. Mr Tzvetkov said that he believed that all of the items in PwC's Net Debt Statement were potentially valid adjustments but, inevitably, some were stronger than others and a degree of discretion had to be exercised. Items 2, 4 and 5 were those on which he insisted there would have to be an agreed adjustment. He did not suggest that those items were non negotiable, and indeed the remainder of the meeting and the subsequent discussions took place on the premise that there remained a negotiation to be had about these items.

(2) Had it been made, it would not have been a representation of fact, but merely a contention or argument made in the course of negotiations, and would reasonably have been understood as such.

(3) Moreover and in any event, Mr Tzvetkov could not reasonably have been understood as doing more than expressing an opinion in the course of the negotiation, rather than making a statement of fact. There was not advanced a case that he misrepresented his state of mind (and I would in any event have found that it would have been an opinion which he honestly and reasonably held). There is for that reason also no warrant for treating it as an actionable misrepresentation. This is not a case like Esso Petroleum v. Mardon [1976] QB 801, upon which the Sellers relied, in which the representor had special expertise or experience on which the representee could reasonably rely so as to treat expressions of opinion as statements of fact.

(4) Had the alleged representation been made and been actionable, the Sellers would not have relied upon it, as I have explained in the context of the alleged misrepresentation about item 1. Ms Ivanova and Ms Dimitrova always believed that items 2, 4 and 5 were not correct and that the adjustment suggested against those items by PwC was not justified.

(5) Had the alleged misrepresentation been made and been an actionable representation of fact, it would not have been made negligently or without an honest belief in its truth. Items 2, 4 and 5 totalled just under BGN 2.35 million, which was less than the adjustment sum which he identified at the meeting as that required by the Buyer, and less than the amount of the adjustment in the settlement ultimately negotiated. He honestly and reasonably believed that items 2, 4 and 5 were valid adjustments.

138. The third misrepresentation alleged is that at the technical meetings between PwC and the Sellers on 15 and 16 December 2009, PwC (by Ms Nalbantova) and/or Mr Georgiev and Mr Rujinov (of the Company) represented, on behalf of the Buyer, that the data for the deferred revenues used by PwC (item 4) had been correctly extracted from the Company's billing system and was accurate save for only insignificant errors. This misrepresentation plea fails for each of the following reasons:

(1) There was no such representation by anyone with actual or ostensible authority to make it on behalf of the Buyer. PwC had no such authority and the context made it clear that anything which may have been said by their representatives in the technical meetings was said on their own behalf to defend their Net Debt Statement. In any event the evidence from the Sellers' witnesses was not that Ms Nalbantova had herself asserted the accuracy of the data used or its source, but that she had invited Mr Georgiev and Mr Rujinov, the IT technicians at the Company, into the meeting to deal with the accuracy and source of the data relied on. The account of the meeting given by the witnesses does not support any representation having been made by Ms Nalbantova that the figures had been extracted correctly. It was clear to all involved that the figures had been extracted not by PwC but rather by Mr Georgiev and Mr Rujinov, and, in the circumstances, it is unsurprising that, rather than Ms Nalbantova giving any confirmation that they had been extracted correctly, she called in the people who had done the extraction to confirm this. Those employees of the Company had no actual or apparent authority to make representations on behalf of the Buyer.

(2) Had any such representation been made on behalf of the Buyer, the Sellers would not have relied on it. Ms Ivanova and Ms Dimitrova did not believe that item 4 of PwC's Net Debt Statement was accurate and correct or that PwC had accurately calculated the figures. On the contrary they continued to believe that PwC had miscalculated the deferred revenues and that their figures were not supported by the Company's systems. They decided to proceed down the route of a negotiated settlement because they did not want to pay for the investigation by an arbitrating accounting firm and take the risk that such firm would reach a figure for the adjustment which was greater than that which they could negotiate in a settlement.

(3) Had the alleged misrepresentation been made it, would not have been made negligently or without an honest belief in its truth. Mr Tzvetkov did honestly and reasonably believe that item 4 was a valid and justified adjustment. There was expert accountancy evidence adduced before me on behalf of the Buyer supporting the correctness of PwC's analysis of item 4. There is no warrant for finding that Ms Nalbantova, Mr Georgiev or Mr Rujinov did not honestly or reasonably believe what they said about item 4 or the data upon which it was based.

139. The fourth misrepresentation was pleaded in the following terms:

"Repeated representations by the Defendant and PwC acting on behalf of and for the Defendant that: (i) PwC's Net Debt statement accurately reflected the underlying accounting information and documents; (ii) the data used by PwC had been accurately extracted from [the Company's] accounting and billing system; (iii) PwC had accurately calculated the figures; and (iv) PwC's Net Debt Statement was accurate and correct. These representations were also false and incorrect."

140. Insofar as this is intended to be founded on what Mr Tzvetkov said at the 14 December meeting or what happened at the PwC technical meetings on 15 and 16 December 2009 it fails for the reasons set out in relation to the first three alleged misrepresentations. Insofar as this plea is intended to apply to anything other than what is alleged in the first three misrepresentations, it contains no clue as to who is said to have made the representations and when. It was not pursued in this form at the trial, although Mr Karia QC relied upon aspects of it to justify a newly formulated misrepresentation plea which I address below.
141. The fifth alleged misrepresentation was pleaded as follows:

"The implicit representation, to be implied from all the circumstances, including inter alia the statements made and the reassurances given by PwC and the Defendant to the Claimant in the context of the requirements of the SPA (particularly, clause 3.4), that PwC had been supplied with all the necessary accounting information and data needed for it to assess [the Company's] Net Debt correctly and accurately."

142. This allegation is embarrassingly vague and was not put to Mr Tzvetkov in cross-examination. There was no such representation made by or on behalf of the Buyer.
143. During the course of the hearing the misrepresentation claim was reformulated beyond the pleaded case. In its final form it included a new set of alleged misrepresentations. The reformulated case was that by sending the PwC Net Debt Statement on 1 December, and/or by Mr Tzvetkov saying at the meeting of 14 December that he trusted PwC and that PwC had done its job professionally, the Buyer thereby represented to the Sellers that:

(1) PwC's adjustment accurately reflected the underlying accounting information and documents; and

(2) in carrying out its review, PwC had acted as an expert (as required by clause 3.4.1 of the 2009 SPA); and

(3) PwC had accurately calculated the figures contained in its adjustment; and

(4) The figures represented PwC's genuine, independently held opinion.

144. Of these, (2) and (4) were not pleaded; nor were they identified in the Sellers' opening, or put to the Buyer's witnesses. It would be unfair to the Buyer to allow them to be advanced. They would in any event have failed. As to (2), the allegation that PwC had not acted as an expert is contrary to the recital in the Net Debt Agreement and is unfounded. The Sellers submitted that the adjustment did not reflect PwC's independently held expert opinion, but was rather a negotiating document produced by PwC in collaboration with the Buyer and, to a very large extent, on the Buyer's instructions. This submission was not borne out by the evidence. The undoubted fact that PwC discussed the contents of the adjustment with the Buyer prior to finalising the PwC Net Debt Statement, and that the latter adopted a large element of the "aggressive" numbers from the draft report, is not inconsistent with the result representing PwC's own independent expert view. The extensive expert evidence adduced before me in respect of each of the items revealed that, at the lowest, it was an opinion which a firm of accountants might reasonably hold on the basis of the contractual exercise which fell to be performed and the information then available. As to (4), the suggestion that PwC did not hold the opinions expressed in the PwC Net Debt Statement, an unparticularised and unpleaded allegation of fraud, is equally unfounded. Still less could it be said that the Buyer did not honestly or reasonably believe that PwC held the opinions which the latter expressed in the PwC Net Debt Statement.
145. The reformulated misreprentations (1) and (3) were not objected to by Mr Toledano QC as outside the pleading. But they were not made out on the evidence. All the Buyer was doing in putting forward PwC's Net Debt Statement on 1 December was purporting to follow the contractual procedure laid down in clause 3.4. It did indeed follow that contractual procedure, as the Sellers themselves agreed it had in the recitals to the Net Debt Agreement. The Buyer did not thereby make any independent representation as to the accuracy of the figures adopted by PwC or the sufficiency or accuracy of the data used to produce them. Subparagraph (4) of this plea recognises that the figures in the Net Debt Statement are in any event matters of opinion rather than objective fact, as was put beyond doubt by the extensive conflicting expert evidence adduced on each side on the individual items and categories. So far as concerns what Mr Tzvetkov said about PwC at the 14 December meeting, Mr Tzvetkov accepted that the minutes were accurate in recording him as saying: "[the Buyer] trusts PwC and the team of Mrs. Nalbantova without any reserves and that they professionally did their job. It was the right of the Sellers to disagree with the results." But this cannot support the misrepresentation case pleaded. In particular:

(1) The PwC Net Debt Statement was PwC's work and contained statements of opinion which Mr Tzvetkov did not represent to be objectively established fact merely by saying that he trusted the PwC team and their professionalism. The context was a negotiation, and the words were used in that context, as was emphasised by Mr Tzvetkov's qualification of his trust in PwC by the expressed recognition that it was the right of the Sellers to disagree with the results. Mr Tzvetkov had not performed the Net Debt review and in saying what he did about PwC he was not making any independent representation that PwC's review was accurate as a matter of objectively verifiable fact, nor that PwC had correctly extracted any relevant data. Nor could he reasonably have been understood as doing so.

(2) In any event had such a representation been made, it would not have been relied upon by Ms Ivanova or Ms Dimitrova, who vehemently disagreed with PwC's figures and said so. Ms Ivanova and Ms Dimitrova believed throughout that PwC's figures were not accurate and were not based on accurate or sufficient data. They did not rely upon Mr Tzvetkov's opinion of them.

(3) Moreover, Mr Tzvetkov honestly and reasonably believed what he said. He honestly and reasonably believed that PwC had used a team which he trusted and had done a professional job.

Breach of clause 5.3

146. Clause 5.3 required the Buyer to provide to the Sellers "full access… to all accounting information and documents on the grounds of which the Net Debt review……..is carried out." The Sellers contended that on a natural reading of the clause, the required access must encompass not only the data extracted and analysed by PwC for the purposes of its review but also access to (a) the whole of the Company's billing system and accounting records; and (b) PwC's methodology and formulae used to calculate the figures; and (c) PwC's analyses and any supporting information underlying their calculations, including the 16 November draft Report.
147. This is not what the clause provides for. It falls to be construed as obligation imposed upon the Buyer, not upon PwC; and keeping in mind that PwC's working papers would belong to PwC, not the Buyer, and would not be within the latter's control. Against that background, it is, as its language suggests, concerned with the specific documents and information to which PwC have had access, and which PwC have taken into account in their review. That is what is meant by "on the grounds of which the …. review…is carried out." Clause 3.4.2 is the clause dealing with the scope of documentation and information to which the parties must afford PwC access. Clause 5.3, by contrast, is concerned with the information and documentation to which PwC has in fact had access. It is concerned with documents and information which form the grounds on which the review is carried out, that is to say the substratum of factual data upon which PwC's analysis is performed. It does not extend to data or extracts of data which were not accessed or considered by PwC. It does not extend to PwC's own documents which contain the analysis itself or to any working drafts or calculations. It does not extend to the detailed workings of PwC's methodology. In those circumstances it is clear that there was no breach of clause 5.3 by the Buyer.
148. In fact the Buyer was content for PwC to provide the analyses which it had used in its draft report, and PwC was prepared to do so on the terms of a release letter which it was reasonable for PwC to require. Had I found that clause 5.3 extended to such documents and information, I would have held that the Sellers' unreasonable refusal to provide the indemnities contained in the release letter would have precluded the allegation of breach.
149. The Buyer advanced a number of further arguments in answer to the claim for breach of clause 5.3, which I shall address in turn. They were that:

(1) by entering into the Net Debt Agreement, the Sellers compromised any claim under clause 5.3; and/or

(2) by entering into the Net Debt Agreement, the Sellers are estopped from alleging any breach of clause 5.3; and/or

(3) by entering into the Net Debt Agreement, the Sellers are estopped from asserting that the Net Debt is anything other than as agreed in that agreement; and/or

(4) if the Buyer was in breach of clause 5.3, and the PwC Net Debt Statement was flawed, this did not cause the Sellers any loss because any such breach did not deprive the Sellers of the ability to contest the PwC Net Debt Statement; on the contrary they did (repeatedly and vigorously) contest it.

Compromise

150. The Sellers submitted that there is no term of the Net Debt Agreement which addresses clause 5.3 or which purports to compromise, or even deal with, any causes of action which the Sellers may have had at that time for the Buyer's breach of clause 5.3; and that it is therefore incorrect to describe or treat that agreement as a compromise agreement affecting the Sellers' causes of action against the Buyer for breach of the 2009 SPA. The Buyer submitted, correctly in my view, that the relevant question is simply what the Net Debt Agreement, construed in its commercial context, would reasonably be understood to mean. The starting point must be the language used by the parties, read against the background of the particular transaction: The Proctor & Gamble Company v Svenska Cellulosa Aktiebolaget SCA [2012] EWCA Civ 1413 at [16].
151. Recitals (3) and (4) record that the Net Debt Agreement was concluded as part of the contractual procedure set out in clauses 3.4 to 3.6 of the 2009 SPA and that the parties were "willing to agree on the Net Debt Amount and on the corresponding adjustment to the Tranche 2 Purchase Price". Clause 2.1 provided that the parties "hereby agree that the Net Debt Amount of the Company is BGN 45,625,990". The background and commercial objective of the Net Debt Agreement was to come to a final and binding agreement on the net debt of the Company and, therefore, the Tranche 2 Purchase Price. Under the procedure set out in clauses 3.4 to 3.6 of the 2009 SPA, this finality was to be achieved either by the parties' agreement (clause 3.6.2(a)) or, in default of such agreement, referral to an arbitrating accounting firm (clause 3.6.2(b)). The parties chose the former route.
152. In these circumstances a reasonable person would have understood the Sellers to have compromised any claim under clause 5.3 of the 2009 SPA, because it would be contrary to the language of the Net Debt Agreement, its background and commercial objective for such a claim to have been preserved. The allegations of breach of clause 5.3 were vigorously and persistently made in the negotiations leading to the Net Debt Agreement. The Net Debt Agreement constituted a compromise agreement as to the amount of net debt in the face of those contested allegations of inadequate access to information and documents. It was plainly intended, in my view, to be a final settlement of the net debt amount notwithstanding those contested allegations of inadequate access to documents and information. If the Sellers could thereafter reopen the net debt adjustment by pursuing a damages claim under clause 5.3, the Net Debt Agreement would not be a final settlement of the net debt or a final and binding adjustment to the Tranche 2 Purchase Price. The Sellers would thereby be permitted to circumvent the contractual mechanism for final determination of the net debt which the parties had put in place in the 2009 SPA. If there remained a dispute between the parties, the dispute should have been referred to an arbitrating accounting firm.
153. The Sellers argued that the rights under clause 5.3 were preserved by clause 15 of the 2009 SPA and clause 4.1 of the Net Debt Agreement. Clause 15 of the 2009 SPA provides that:

"15. Waiver

A delay in exercising, or failure to exercise, any right or remedy under this Agreement does not constitute a waiver of such right or remedy or other rights or remedies nor shall either operate so as to bar the exercise or enforcement of such right or remedy."

Clause 4.1 of the Net Debt Agreement provides that:

"The provisions of Clauses 12 to 16, 19, 20 and 24 to 26 of the [2009 SPA] are hereby incorporated by reference into, and made a part of, this agreement as if fully set forth herein in full, mutatis mutandis."

154. In my judgment these clauses do not assist the Sellers, because the Buyer is not alleging that the Sellers waived any rights under clause 5.3 by a "delay in exercising, or failure to exercise" those rights. The argument is that the Sellers compromised such rights by entering into the Net Debt Agreement.
155. The Sellers also advanced an argument in their Amended Reply that clause 3.2 of the Net Debt Agreement preserves their rights under clause 5.3 of the 2009 SPA. Clause 3.2 of the Net Debt Agreement provides:

"Except as amended in accordance with this agreement, the [2009 SPA] remains unchanged and shall continue in full force and effect, and all references to the [2009 SPA] shall be a reference to such agreement as amended hereby."

156. However, this begs the question of whether on the true construction of the Net Debt Agreement, the Sellers did compromise any claims under clause 5.3 of the 2009 SPA. It does not take the Sellers' argument on the point any further.
157. For these reasons I would accept the Buyer's argument that the Net Debt Agreement compromises any cause of action for breach of clause 5.3.

Promissory estoppel

158. The Buyer's argument on promissory estoppel was closely connected to its argument on compromise. The representation alleged is that by their conduct in entering into the Net Debt Agreement, the Sellers impliedly represented that they would not pursue a claim for breach of clause 5.3, because the Net Debt Agreement was inconsistent with the subsequent bringing of a claim for breach of clause 5.3. Whilst recognising that a promissory estoppel can be founded on a representation to be implied from conduct, I do not think that this analysis assists the Buyer. If the Net Debt Agreement was a compromise of any claims for breach of clause 5.3 because it was inconsistent with their continued existence, as I have found, it constituted a contractual agreement that the Sellers would not and could not pursue such claims. If, on the other hand, I had considered that the Net Debt Agreement did not constitute an agreement to compromise such claims, there would have been no representation by the Sellers by their conduct in entering into the agreement that they would not pursue such claims. In either event, the doctrine of promissory estoppel would have no part to play.

Contractual estoppel, estoppel by convention, estoppel by representation of fact

159. These arguments focus on the amount of the net debt itself. The Buyer contended, correctly in my view, that it is a necessary part of the Sellers' claim for damages for breach of clause 5.3 that, as a matter of fact, the Company's net debt was less than that agreed in the Net Debt Agreement. This is because the Sellers' claim depended upon their argument that, had they been provided with the relevant information and documentation, they would not have entered into the Net Debt Agreement; they would either have been able to persuade the Buyer and PwC of the latter's erroneous adjustments, or would have had the matter determined by an arbitration accounting firm; and that in either event the net debt would have been less than the amount fixed in the Net Debt Agreement. Unless the true net debt, as agreed between the parties or determined by the arbitrating accounting firm (in each case on the hypothesis of provision of further information/documents), was less than that to which the Sellers agreed in the Net Debt Agreement, there could be no loss suffered. The Buyer contended that the Sellers are estopped from alleging, as a matter of fact, that the Company's net debt is anything other than the figure agreed in the Net Debt Agreement because in that agreement both parties agreed that the net debt of the Company was BGN 45,625,990 as a matter of fact.
160. The plea based on contractual estoppel is, in my view, well founded. Clause 2.1 of the Net Debt Agreement provided that: "The Sellers and the Buyer hereby agree that the Net Debt Amount of the Company is BGN 45,625,990". This is a contractual agreement as to the factual position. Where parties settle in a contractual document a disagreement as to an existing state of affairs in order to establish a clear basis for the contract itself and for its subsequent performance, neither party can subsequently deny the existence of the facts and matters upon which they have agreed: Peekay Intermark Ltd v Australia and New Zealand Banking Group Ltd [2006] EWCA Civ 386 at [56], Springwell Navigation Corporation v JP Morgan Chase Bank [2010] EWCA Civ 122. For such an estoppel it is not necessary to show that it would be unconscionable for a party to resile from the agreed position (see Springwell at [177]-[178]), but in any event I would regard it as unjust that the Sellers should be able to resile from the express agreement reached in the Net Debt Agreement as to the amount of the net debt, which was a freely negotiated commercial compromise. It follows that the Sellers are contractually estopped from denying that the Company's Net Debt Amount was anything other than BGN 45,625,990, which is fatal to their claim for breach of clause 5.3 of the 2009 SPA.
161. Nothing is added to the Buyer's case by the alternative formulation of this point under an analysis of estoppel by convention or estoppel by representation.

Causation

162. The Sellers' case was that, as a consequence of the Buyer's alleged breach of clause 5.3 of the 2009 SPA, they were unaware that the PwC Net Debt Statement was incorrect; and that as a consequence, they were deprived of the opportunity to challenge the PwC Net Debt Statement, at least so far as concerned items 2, 4 and 5. This allegation of causation fails on the facts. The Sellers repeatedly challenged the accuracy of the PwC Net Debt Statement and at the time they entered into the Net Debt Agreement, Ms Ivanova and Ms Dimitrova still did not believe that the PwC adjustment was fair or accurate in respect of those or the other items. They nevertheless made a conscious choice, for commercial reasons, to negotiate the net debt with the Buyer instead of either proceeding to instruct an arbitrating accounting firm or making any further complaints about their lack of access to information. Accordingly, had the Buyer been in breach of clause 5.3 of the 2009 SPA in denying access to information or documents, such lack of access would not have deprived the Sellers of the ability to challenge the PwC Net Debt Statement. The claim for breach of clause 5.3 fails also on causation.
163. For each of the reasons identified above (no breach, compromise, contractual estoppel and causation), the claim for breach of clause 5.3 fails.

Breach of clause 3.4.2

164. The Sellers' case is that the Buyer was in breach of clause 3.4.2 of the 2009 SPA ("In so far as they are able, the parties shall each provide, and procure that the Company provides, PwC with all information relating to the Company which PwC reasonably requires, including but not limited to all relevant breakdowns and analytical information related to the Net Debt.") because Mr Tzvetkov expressly instructed employees of the Company to provide PwC with only such documents as they specifically requested; and that given the alleged errors in PwC's Adjustment, it is reasonable to assume that other important information and documentation was not provided, without which PwC could not adequately perform the review required of it by the 2009 SPA.
165. This case also fails for each of the following reasons:

(1) There was no breach. The clause only requires provision of the information requested by PwC. That is what was meant by information "reasonably required" by PwC. There is no suggestion that PwC were denied access to any documents or information it considered necessary for its review. The clause does place the burden of judging what material PwC might need for the purposes of its review on the Buyer rather than PwC.

(2) Any claim for breach of clause 3.4.2 was compromised by entering into the Net Debt Agreement, for the reasons identified in relation to claim for breach of clause 5.3. Indeed the position is even clearer in respect of clause 3.4.2 because the Net Debt Agreement records in Recital 3 that the procedure required by clause 3.4 has been followed.

(3) The Buyer is contractually estopped from challenging the net debt amount, which is fatal to the claim for the same reasons as explained above in relation to the claim for breach of clause 5.3.

Loss

166. In the further alternative, the Buyer took issue with the Sellers' case as to the true amount of the net debt, and the extent to which there would have been an adjustment to the Sellers' Net Debt Amount had the parties not entered into the Net Debt Agreement. I received a considerable body of evidence on this topic in relation to many of the 22 individual items of the adjustment, including expert accounting evidence on both sides. On my findings this question does not arise, and it would be inappropriate to burden this already overlengthy judgment with a detailed analysis of the myriad sub issues debated. It is sufficient to state my conclusions that the Sellers failed to persuade me that PwC would itself have determined the adjustment to be within €100,000 of the Sellers Net Debt Amount so as to avoid the need for an arbitration by an accounting firm; or that the matters referred to arbitration by an accounting firm would have been limited to items 1, 2, 4 and 5; or that the true net debt was such that it would have involved an adjustment of less than BGN 2.75 million. Accordingly had there been a claim in misrepresentation or for damages for breach of clause 5.3 or 3.4.2, the Sellers would have failed to prove any loss.
167. For all these reasons the Sellers' net debt claim fails.

Conclusion

168. The Sellers are entitled to payment of the second tranche of the purchase price, as reduced by the Net Debt Agreement, in the sum of €4,194,000. The Buyer's counterclaims fall to be dismissed.