England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE, BUSINESS AND PROPERTY COURTS OF ENGLAND AND WALES, BUSINESS LIST (ChD)
Stephen Jourdan KC sitting as a Deputy High Court Judge
[2022] EWHC 1597 (ChD)

		Royal Courts of Justice Strand, London, WC2A 2LL
		14 March 2023

This judgment was handed down by the Court remotely by circulation to the parties' representatives by email and release to The National Archives. The date and time for hand-down is deemed to be 10:30 on 14 March 2023.

Lord Justice Arnold:

Introduction

1. This is an appeal by the Claimant ("Manolete") against an order of Stephen Jourdan KC sitting as a Deputy High Court Judge dated 24 June 2022 dismissing Manolete's claim against the First Defendant ("Mr Dalal") for misappropriation of money from Bolton Poultry Ltd ("the Company") for the reasons given in the judge's judgment of the same date [2022] EWHC 1597 (Ch). The judge also dismissed claims by Manolete against Mr Dalal's son Sajid and daughter-in-law Anisha and the estate of his late wife Johra, but there is no appeal by Manolete against those parts of his order, and those claims can be ignored for the purposes of the appeal.

The claim

2. The Company was incorporated on 18 April 2005. Until it went into creditors voluntary liquidation on 30 January 2015, it carried on business in the Bolton area as a slaughterer and wholesaler of halal chicken meat. It supplied restaurants and retailers with (i) whole chickens and chicken parts which the Company had slaughtered and (ii) fresh and frozen whole chickens and chicken parts which the Company had itself bought in from suppliers. Mr Dalal was a director of the Company throughout this period. Sajid, Anisha and Johra were directors until 20 June 2008, but not thereafter. Johra passed away in May 2018.
3. Manolete claims as assignee from the Company and its liquidator of claims against Mr Dalal. In summary, the claim is that: (i) over the period of the Company's trading life its turnover was considerably greater than the overall total of £30,716,557 that appeared in its annual accounts and had been reported to HMRC in its corporation tax returns; and (ii) Mr Dalal, as director of the Company, is liable to account for the additional profits the Company made.
4. As is explained in more detail in the judge's judgment, the genesis of the claim was an investigation conducted by Her Majesty's Revenue and Customs ("HMRC") between 2011 and 2015 into the Company's corporation tax return for the year ending 30 September 2009. HMRC had been provided with information obtained by officers of the Meat Hygiene Service, and subsequently the Food Standards Agency ("FSA"), who, in accordance with their prescribed statutory functions, attended at the Company's premises and (and amongst other things) recorded the quantities and weights of the chickens entering the premises.
5. In January 2013 the Company's then accountants SCB sent HMRC completed Outline Disclosure forms on behalf of each of the Defendants. Mr Dalal's form acknowledged that there appeared to be understatements in the Company's accounts for the years ending 30 September 2006 and 2007, but said it was "less clear" whether there were any understatements in subsequent years.
6. The investigation resulted in a report referred to as the Business Economics Exercise ("the BEE") which HMRC sent SCB in March 2014. The BEE estimated the true level of the Company's sales for the year ending 30 September 2009. It was prepared on two bases: method 1 used a sale price of £2.75 per kilogram of chicken derived from "HMRC 3^rd party sources". Method 2 analysed the actual sale prices of 43 items across five of the Company's invoices to three customers in 2009, yielding an average price of £2.62 per kg of chicken. Method 1 resulted in an estimate of total sales of £6,179,200. Method 2 resulted in an estimate of £5,887,092. Both figures were significantly greater than the figure of £2,695,644 recorded in the Company's accounts for that year (and even more so if sales of livestock and "purchases for own use" recorded in the accounts of £281,244 were deducted from the latter figure).
7. In August 2014 the Company instructed new accountants, AMS, to respond to the BEE. In late November 2014 AMS wrote to HMRC accepting that the Company's record keeping was "far from perfect" and that there was "some evidence that the declared sales figure was incorrect". The letter went on:

"Having discussed your figures with our client, we believe that the level of additional sales should be closer to 20% of the additions proposed in your letter."

8. In June 2015 HMRC served corporation tax assessments on the Company's liquidator which were revised in March 2016. Using method 2 in the BEE, HMRC determined that there had been total sales of £5,887,092 in the year ending 30 September 2009. Applying the presumption of continuity, HMRC assumed that the Company had made the same level of sales in every year it had traded. HMRC deducted the sales reported in the Company's accounts. This produced a total figure of additional sales of about £36.6 million. HMRC treated the AMS letter as an admission that the figure was at least 20% of that total, and thus calculated the Company's liability to corporation tax on the basis that it had made additional sales of at least £7,123,644 over the 10-year period.
9. In September 2015 AMS produced a report ("the AMS Report") concluding that sales had been under-reported, but only to the extent of £849,278, and that costs had also been under-reported, leaving an under-reported trading profit of £61,249. As explained in more detail below, the AMS Report included analyses of both sales invoices and purchase invoices.
10. The liquidator assigned all rights to pursue claims against the Defendants in July 2019. In November 2019 Manolete commenced these proceedings. As explained in more detail below, Manolete adopted HMRC's calculations in its Particulars of Claim and claimed a total of £7,123,644 in respect of the claim presently under consideration.
11. There is no dispute as to the legal basis for Manolete's claim, but Mr Dalal denies that the Company made sales additional to those reported in its accounts and corporation tax returns. This is a pure question of fact turning upon the evidence before the judge.
12. The documentary evidence available at trial was significantly incomplete for at least three reasons. First, although the FSA records for each year were available, the underlying documentation was not. Secondly, HMRC had destroyed most of the documents produced during the course of the investigation. Thirdly, although it appears that AMS had considerable access to the Company's sales documents for the purposes of producing the AMS Report, none of the documents used by AMS in drawing up the Report were delivered to the liquidator or disclosed by the Defendants in the proceedings. I shall return to this point below.

The judge's judgment

13. Having set out the factual background, the judge identified the witnesses and gave his assessment of them. Manolete called the liquidator, the officer in HMRC's Fraud Investigation Service who had carried out the investigation into the Company and a representative of the FSA. All three were described by the judge as careful witnesses. The Defendants called Mr Dalal, Sajid and Anisha (but no representative of either SCB or AMS). The judge's assessment of their evidence was as follows:

"59. [Mr Dalal] gave evidence from his home in India, via an interpreter. He is 84 years old and suffers from a number of ailments. He said in his witness statement, and again at the beginning of his evidence, that his memory was defective. He struggled to understand many of the questions, and his answers to many of them were that he could not remember the relevant events.

60. Sajid was a very nervous witness, who struggled to understand some of the questions he was asked.

61. Anisha was an intelligent and articulate witness, who listened carefully to the questions she was asked and gave clear answers to them.

62. Insofar as the evidence of the witnesses was challenged, I will evaluate it by reference to the uncontested facts, the documentary material I was taken to in the course of the trial and the inherent probabilities. I do not think any of the witnesses were giving evidence they believed to be false, but human memories are fallible, and the evidence of a witness may be incorrect even if they believe it to be accurate."

14. The judge started his consideration of the present claim by observing at [90]:

"The claim is, therefore, based firmly on the HMRC tax assessments, which in turn were based on the second method used in the HMRC Business Economics Exercise. The issue is whether that method is sound and, if so, whether it produces the conclusion that there were Additional Sales Receipts in the year ending 30 September 2009. If there were, then it will then be necessary to consider the position in respect of the other trading years. As to that, HMRC applied the presumption of continuity, but [the liquidator] produced a calculation which applied the logic of the Business Economics Exercise to the other trading years, which seems to me preferable. However, first it is necessary for me to determine if the evidence establishes that there were Additional Sales Receipts in the year ending 30 September 2009."

15. The judge then discussed some general considerations bearing upon this question, only some of which it is necessary to note for present purposes. First, the judge recorded at [94] that it was common ground between the parties that it was unlikely that the Company had incurred unreported costs, and the issue was simply whether there were additional sales to those it had reported.
16. Secondly, the judge said this at [103] about Mr Dalal's Outline Disclosure form:

"When asked about this, Ebrahim said that he had studied at school in Gujarati. He could speak English when he was in the UK, but he could not read or write English. He was, therefore, dependent on his accountant to complete forms on his behalf. He could not remember the details of what had happened at the time, but: 'What I used to understand is that we keep the accountant and he will do his job and he takes on the responsibility, so I was more concerned or concentrating on my business and doing all the running around.' I accept that evidence which strikes me as being inherently plausible."

17. Thirdly, the judge recorded at [108] that neither side relied upon the reasoning or conclusions of the AMS Report, and concluded that he could not place much weight upon the 20% figure in the AMS letter of November 2014 either. Nevertheless, he said at [109]:

"What I do think the Outline Disclosure and the AMS letter establish is that two different accountants thought that there were deficiencies in the Company's systems for recording sales. If the systems for recording sales had been robust and reliable, one would not find [SCB] filling in the Outline Disclosure form saying that there probably had been some understatement of sales, and AMS saying that 'our client's record keeping was far from perfect'. Accordingly, it is plausible that there may have been unreported sales. I do not think the inadequacies of the record keeping justifies the conclusion that there were unreported sales, but it means that there could well have been."

18. The judge also noted at [112] that the AMS letter was written only two months before the Company went into liquidation, suggesting that nothing had changed about the recording of sales after the HMRC investigation had begun. The judge went on to say at [113] that the fact that parts of the AMS Report might be unreliable did not mean that all of the information in it was unreliable.
19. Fourthly, the judge concluded this section of his judgment as follows:

"114. The trial bundle includes only relatively few documents recording sales and purchases by the Company. Mr Shaw said that this was the fault of the directors, who should have delivered the Company's books and records to the liquidator. He argued that the Defendants are not able to rely on the inability of the books and records which have not been provided to make good what they say otherwise is their position.

115. Given my determination that [Mr Dalal] was the sole director of the Company after 2008, this submission could only be relevant to evaluating his case. However, I do not think it is a valid point even in his case. The Claimant's pleaded case relied on the HMRC Business Economics Exercise, which involved analysing the sales in the year ending 30 September 2009, and then extrapolating the conclusions from that analysis to the remainder of the Company's trading period. The documents on which HMRC relied were destroyed by them. The Claimant made no complaint during the course of the litigation that there had been inadequate disclosure by [Mr Dalal] of books and records from other years. Nor was it suggested to [Mr Dalal], Sajid or Anisha in cross-examination that they had documents which would or might cast light on the issue which they had failed to disclose.

116. Overall, my conclusion from the general points discussed above is that the Claimant's case must stand or fall on an analysis of the evidence which supports or contradicts the Business Economics Exercise calculation.

117. It will be apparent from the discussion below that the evidence on a number of aspects of the calculation was extremely thin, both on the Claimant's side and on the Defendants' side. There is substantial uncertainty about a number of aspects of the calculation. I have considered whether the evidence is so thin that I should simply determine that it is insufficient for me to make any rational decision, in which case I would have to conclude that the Claimant has failed to prove its case: see Verlander v Devon Waste Management Ltd [2007] EWCA Civ 825 at [24]. In the end, I have decided that there is, just, enough information for me to make a very rough estimate of the likely level of sales in the year ending 30 September 2009 which can be compared to the figure for sales in the Company's accounts for that year of £2,695,644."

20. The judge then carried out a careful and detailed analysis to calculate the total sales of chickens and chicken parts not merely in the year ending 30 September 2009, but across whole of the Company's trading life. This consisted of a review of the following elements: (i) the quantities of chickens slaughtered or cut up by the Company; (ii) the average weight of the birds; (iii) the proportion of dead or diseased birds that could not be sold; (iii) the respective proportions of birds that were sold whole and in parts; (v) the wastage incurred in the slaughtering and cutting processes; (vi) the prices at which the Company sold the chickens; and (vii) the level of bad debts.

Quantities of chickens brought into the Company's premises for slaughter or cutting

21. There were two separate aspects to the Company's trade here. The majority of the Company's turnover derived from chickens that it slaughtered. A small proportion consisted of birds it bought in that had already been slaughtered and which the Company cut up. This second element did not form part of the BEE, but nevertheless Manolete relied upon it at trial without objection from the Defendants.
22. The source of the evidence as to the quantity of chickens entering the Company's premises for slaughter was the FSA records, which recorded both quantities and weights of birds. The judge rejected the Defendants' case that the quantities recorded by the FSA represented anticipated (as opposed to actual) throughput (judgment at [123]). The judge found that "the only reliable basis" for carrying out the calculation was that adopted by HMRC, which was to assume that the quantity of birds purchased was accurately recorded by the FSA ([133], [137]). The total number recorded by the FSA in the period from April 2005 to January 2015 was 13,834,436 ([124]).
23. As to the quantity of meat brought into the Company's premises for cutting, the source of the evidence was the Company's own returns to the FSA (as opposed to information recorded by FSA officers). On this basis the judge found that the quantity was 47,000 kg in the year ending 30 September 2009 ([178]) and 2,451,000 kg across the whole trading period (Appendix 2).

Weight of chickens

24. The FSA returns recorded "nearly all" of the birds being delivered to the Company's premises as being in one of the categories of less than 2 kg ([124]). In the BEE, HMRC used an average weight of 2 kg. There was evidence from three sets of invoices from the Company's suppliers that chickens purchased had a greater weight than 2 kg ([127]), including an analysis in the AMS Report of purchase invoices between December 2005 and June 2012 which showed an average weight of 2.61 kg per bird. There was also some evidence in the BEE of the delivery of baby birds and hens having a weight of less than 2 kg ([144-145]). The judge found that "the only reliable basis" for carrying out the calculation was that adopted by HMRC, which was to assume that the average weight of a live bird was 2 kg ([133], [137], [145]). The average of 2 kg was a "very rough and ready figure", which took into account the evidence that some birds would have weighed less and some more ([145]).

Dead or diseased birds

25. There was evidence derived from invoices from one of the Company's suppliers showing that about 2.3% of birds would be dead on arrival or unfit for consumption due to disease, whereas Mr Dalal's evidence was that the figure was sometimes higher. The judge found that an average of 5% was appropriate ([142]).

Proportions sold whole and in parts

26. Counsel for Mr Dalal presented an analysis showing that 33% of slaughtered birds were sold whole and 67% were cut into parts for sale, which the judge accepted ([167]).

Wastage

27. The judge recorded that Mr Dalal's evidence was that the slaughtering and processing of meat would give rise to significant wastage and that approximately 60% of a live bird's weight would be saleable (i.e. a wastage proportion of 40%) ([152]). The judge found that there would be different amounts of wastage between chickens sold whole and chickens cut up into parts – the latter process involving further quantities of unsaleable remains. In addition to Mr Dalal's evidence, the judge considered various other sources of evidence. Having done so, the judge found that the birds were sold whole at 75% of the weight of a live bird and sold cut up into parts at 50% of the weight of a live bird, giving an overall wastage figure once the proportions of whole birds and parts were taken into account of 42%, which compared reasonably with Mr Dalal's estimate of 40% total wastage ([168]).

Sale prices

28. There were two sources of evidence as to the sale prices:

i) The five invoices that were relied on by HMRC in the BEE showed an average sale price of £2.67 per kg. The Defendants pointed out that the calculation in the BEE was erroneous and the correct figure was £2.48 per kg ([170]). The judge said that, if HMRC's method 2 was the only evidence as to average selling price, he would have regarded it as too unreliable to base any conclusions on. It was based on a sample of 0.25% of the Company's sales in that year, which could not be taken to be a reliable method for determining the average sale price ([170]).

ii) 182 invoices from the Company to a customer called Watan Cash and Carry covering the period August 2011 – August 2013 that were in the trial bundles (I will call these "the Watan invoices"). These totalled £135,556, which was 3.7% of the Company's reported total turnover in 2011-12 ([172]). Counsel for Mr Dalal's analysis of these invoices showed consistently the same price for the sale of whole chickens of £1.75 per kg and an average sale price for chicken parts of £2.37 per kg [(173]-[174]). The judge commented that this was "not a very reliable guide", but nevertheless it was "the most reliable source of information" and "the best evidence available" ([172]), and he therefore adopted these figures. That gave total estimated sales from slaughtered birds before allowing for rejected birds and bad debts of £2,670,025 in the year ending 30 September 2009 after allowing for dead and diseased birds and for wastage ([175], [181] and Appendix 1).

Rejected birds and bad debts

29. The parties agreed, and the judge adopted, a figure of 2% for birds rejected by customers and bad debts which AMS had used ([176]).

Overall

30. Deducting 2% from the total estimated sales for the year ending 30 September 2009 resulted in a final figure of £2,616,625, or in round numbers about £2.6 million ([175], [181] and Appendix 1). The judge said he had "very little confidence in the accuracy of that figure", but it was "the best I can do based on the evidence" ([182]). This was less than the figure reported in the accounts of £2,695,644, or in round numbers about £2.7 million ([181] and Appendix 1). Thus the evidence did not establish that there were any additional sales in the year ending 30 September 2009.
31. As the judge recorded, however, counsel for Manolete presented an analysis of the figures over the trading life of the Company using the figures set out above which the judge reproduced in Appendix 2 to his judgment and which I reproduce in the Appendix to this judgment. This exercise produced an estimate of total sales of £34,455,432 against reported sales of £30,716,557, a difference of £3,738,875 (12.2%).
32. The judge dealt with this analysis, and counsel's arguments based upon it, as follows:

"183. In closing, Mr Shaw also presented calculations based on the number of birds slaughtered over the whole of the trading life of the Company. Applying the assessments I have made above, this produces a figure which is about 12% higher than the total reported sales of the Company over its trading life: see Appendix 2. Give[n] the very substantial uncertainty in the accuracy of the figures, I do not think that this is a sufficient discrepancy to prove that there were Additional Sales Receipts. It is possible that there were, but in my judgment the Claimant has not proved that it is more likely than not that there were Additional Sales Receipts.

…

185. The assessment of the likely level of sales which appears in Appendix 2 is the product of a calculation involving a substantial number of variables. In each case, the evidence which I have relied on to make my estimate is poor and my confidence in the accuracy of those estimates is low. This means that there is, inevitably, a substantial margin of error. For example, the estimate of average selling prices per kilo I have used is based on a very small sample from a period of about 2 years out of the total trading life of the Company. A different sample might well produce a different average selling price which would lead to a different output from the calculation. In his closing submissions, Mr Shaw referred to some invoices from 2011 which showed an average selling price of chicken parts of £2.69 per kg. Ms Fisher's analysis referred to above covering some invoices from 2011-2013 produced an average of £2.37 per kg, about 12% lower than Mr Shaw's figure.

186. The conclusion of the Business Economics Exercise was that the Company's sales in the year ending 30 September 2009 were more than 200% of the amount recorded in the accounts. If I had arrived at a similar assessment, then despite the uncertainties in the estimates and the need to allow a margin of error I would have been persuaded that there probably were Additional Sales Receipts, and would have endeavoured to make the best estimate that I could of what they were. As it is, my estimates produce a figure which is only 112% of the reported sales. That is not enough of a discrepancy to justify the conclusion that there probably were Additional Sales Receipts, rather than the discrepancy being due to errors in the accuracy of estimates made on the basis of poor quality evidence. It is not necessary to define what that margin of error is which ought to be applied to the calculation in Appendix 2, but it is certainly higher than 12%.

187. Mr Shaw referred to In Re A (Children) (Care Proceedings: Burden of Proof) [2018] EWCA Civ 1718, in which King LJ gave guidance on the correct approach to fact finding and the application of the burden of proof …. She summarised the guidance at paragraph 58 as follows:

'i) Judges will decide a case on the burden of proof alone only when driven to it and where no other course is open to him given the unsatisfactory state of the evidence.

ii) Consideration of such a case necessarily involves looking at the whole picture, including what gaps there are in the evidence, whether the individual factors relied upon are in themselves properly established, what factors may point away from the suggested explanation and what other explanation might fit the circumstances.

iii) The court arrives at its conclusion by considering whether on an overall assessment of the evidence (i.e. on a preponderance of the evidence) the case for believing that the suggested event happened is more compelling than the case for not reaching that belief (which is not necessarily the same as believing positively that it did not happen) and not by reference to percentage possibilities or probabilities.'

188. As to the first point, I have been driven to my conclusion by the unsatisfactory state of the evidence. If there had been better evidence, I would be much more confident in the accuracy of the estimates of the variables which I have made, and the margin of error would be smaller.

189. As to the second point, I have endeavoured to look the whole picture, and all relevant factors, as I hope is apparent from my judgment above.

190. As to the third point, I have not attempted a probability calculation. Rather, I have undertaken a calculation using estimates of a number of variables. In respect of each of the variables I have done my best on the limited evidence available, but there is substantial uncertainty. I have little confidence in the accuracy of the estimates and therefore little confidence in the accuracy of the output of the calculation. The discrepancy between the output in the calculation and the reported sales figures in the Company's accounts could be due to there having been Additional Sales Receipts, or could be due to errors in the estimates. If the discrepancy was very large then I would be satisfied that it was more likely than not to be due to there having been Additional Sales Receipts. However, it is not very large; it is about 12%. On Mr Shaw's submission, even if the discrepancy was only 1% I would be bound to determine that there had been Additional Sales Receipts. It seems to me obvious that this is wrong."

Grounds of appeal

33. Manolete contends that, based on the Appendix 2 calculation, the judge should have found its claim proved on the balance of probabilities to the extent of £3,738,875. It appeals on four grounds. Ground 1 is that the judge's findings impermissibly went beyond Mr Dalal's pleaded case, evidence and submissions at trial. Ground 2 is that the judge's reasoning contained identifiable flaws or gaps in logic and/or was inconsistent in its evaluation of the evidence. Ground 3 is that the judge adopted the wrong approach to the application of the burden of proof. Ground 4 is that the judge failed to take into account the limits on the evidence due, Manolete contends, to Mr Dalal's default in failing to provide books and records of the Company to the liquidator or on disclosure. For reasons that will appear I will consider these grounds in a slightly different order.

Ground 4

34. Ground 4 was only faintly pursued at the hearing, and rightly so. Manolete relies on the principle stated by Arden LJ in Re Mumtaz Properties Ltd [2011] EWCA Civ 610, [2012] 2 BCLC 109 at [17] that, where the books and records of a company are inadequate due to the failure of the directors properly to maintain them and/or to deliver them up to the liquidator (or other office holder), the directors will not be able to claim that, had proper records been maintained and delivered up, they would have demonstrated that the directors were not liable for the sums claimed. This principle does not assist Manolete. The judge accepted that the Company's record keeping had been less than perfect, and that it was therefore possible that sales had been underdeclared. He went on to make findings based on the evidence that was available. As he explained, it was not suggested to the Defendants that they had failed to disclose relevant documents. Nor did the Defendants argue that better records would have disproved Manolete's claim.

Ground 1

35. Although ground 1 was not strongly pressed, it is necessary to deal with it in some detail since it provides important context for ground 2, which formed the main plank of counsel for Manolete's argument.
36. Manolete's Particulars of Claim (which were not settled by counsel who appeared for Manolete at trial and in this Court) pleaded the claim in rather a curious manner. In paragraph 10 Manolete recited HMRC's revised assessments. In paragraph 11 it said that the basis of the assessments, which were raised following an investigation by HMRC, was set out in documents produced by HMRC, copies of which had been provided to the Defendants (i.e. the assessments and the BEE). In paragraphs 12.1 to 12.3 it briefly summarised the calculation in the BEE and the way in which it had been extrapolated by HMRC. In paragraphs 12.4 and 12.5 it recorded HMRC's acceptance of the assertion in the AMS letter of November 2014 that the true level of additional sales was more like 20% of HMRC's figure. In paragraph 13 it asserted that, on the basis of the assessments, the Company had made no less than £7,123,644 of sales additional to those shown in the Company's accounts. Thus the claim was predicated upon the HMRC calculations, and hence upon the BEE. It did not set out the constituent elements of those calculations, however.
37. In paragraph 11 of his Defence Mr Dalal pleaded to paragraphs 10 to 12 of the Particulars of Claim as follows:

"The accuracy of those Assessments is denied. In particular the HMRC employed underlying methods, estimations and assumptions which were fundamentally flawed in that: …"

38. Four matters were then set out: (i) HMRC had failed to allow for wastage; (ii) HMRC had failed to allow for dead and diseased birds; (iii) HMRC's figures were based on a retail model (this was a reference to method 1 in the BEE), whereas the Company was a wholesaler: and (iv) the FSA records represented anticipated rather than actual numbers of birds slaughtered.
39. Counsel for Manolete pointed out that there was no specific challenge to the sale prices used in method 2 of the BEE. That is true, but there was a general denial of the accuracy of the HMRC figures.
40. Perhaps more importantly, paragraph 12 of the Defence noted that there was no reference in Particulars of Claim to the AMS Report in which AMS, it was pleaded:

"a. Carried out an in depth analysis of the Company so as to assess the correct level of sales from incorporation to 30 September 2014;

b. Investigated the correct level of turnover by sampling quarterly sales and purchase invoices, in parallel with an analysis of the average cost per kg for each bird slaughtered;

c. Concluded that the Company's turnover had been understated by £849,278, resulting in understated gross profit of £61,249 (which is explained by way of withdrawals from the First Defendant's director's loan account)."

41. At that stage, therefore, Mr Dalal was relying upon the accuracy of the figures in the AMS Report. This is significant for three reasons. First, whereas the HMRC calculations were based on extrapolations from the figures for the year ending 30 September 2009, the AMS Report analysed data for almost the whole of the trading life of the Company. Secondly, the AMS Report included data as to the weights of birds slaughtered which the judge considered. Thirdly, the AMS Report also included data for the prices charged by the Company ranging from a low of 0.25 per kg (for baby birds) to a high of £3.75 per kg (for breast meat). The data does not enable average figures to be calculated which can be compared with the other evidence in the case. As one might expect, the prices given for the end of the period are generally appreciably higher than those at the beginning.
42. In paragraph 3 of its Reply Manolete took issue with each of the four matters pleaded in paragraph 11 of the Defence. In paragraph 4 Manolete took issue with the reliability of the AMS Report. In particular, Manolete pleaded that (i) AMS had failed to take into account the slaughtered chickens which the Company had bought in and (ii) AMS' calculations were based upon an average weight of 2 kg whereas the invoices sampled by AMS showed an average weight of 2.6 kg. In paragraph 5 Manolete pleaded that calculating the sales of the Company using (i) the quantity of chickens recorded by the FSA, (ii) an average weight of 2 kg and (iii) an average price of £2.75 kg (i.e. the method 1 price) gave a total of £57,808,207 as opposed to the total turnover of £30,069,823 recorded in the Company's accounts.
43. In paragraphs 16 to 18 of his witness statement for trial Mr Dalal repeated the four criticisms of the accuracy of the HMRC figures set out in his Defence, but did not make any other criticism of those figures. Nor did Sajid or Anisha.
44. In opening submissions at trial counsel for Manolete relied upon the HMRC calculations, including both method 1 and method 2. In closing submissions counsel for Manolete abandoned reliance upon method 1. He did not merely rely upon the HMRC figures, however, but also made submissions as to what the evidence showed in respect of each of the elements in the calculation discussed above. In relation to prices, counsel for Manolete presented two analyses of the available invoices. The first was that the five invoices used in method 2 of the BEE showed that whole birds were sold for £1.80 per kg while chicken parts were sold for an average of £2.51 per kg. The second was that the first 10 of the Watan invoices showed a uniform price of £1.75 per kg for whole chickens and an average price of £2.69 kg per kg for parts (this is the analysis referred to by the judge in [185]).
45. In closing submissions counsel for Mr Dalal (Ms Fisher, Mr Leiper KC not having appeared below) also analysed each of the elements of the calculation. She submitted that the price of £2.75 used in method 1 should be rejected as unevidenced and that the price of £2.62 used in method 2 was unreliable because it was based on only five invoices over a short period time. (Although she also relied upon evidence from Sajid that the price could be as low as 80p per kg, that evidence was not in his trial witness statement and in any event his oral evidence was that he did not know the sale prices.) In response to counsel for Manolete's analysis of the first 10 of the Watan invoices, she presented an analysis of all 182. This confirmed that the consistent price for whole chickens was £1.75 per kg and showed that the average price for parts was £2.37 per kg (this was the analysis referred to by the judge at [172]-[-175] and [185]). She nevertheless made the overarching submission that Manolete could not prove its case because there was insufficient evidence of underdeclarations of sales given the margins of error in the calculations.
46. Against that background, counsel for Manolete accepted during the course of argument that the judge's findings had not gone beyond Mr Dalal's pleaded case. As discussed above, Mr Dalal had pleaded a general denial of the accuracy of the HMRC figures. It was therefore open to Mr Dalal to contest each and every element in the calculations, and in any event to contest the reliability of the calculations for the purpose of proving that the Company had significantly underdeclared its sales. Furthermore, it was only in its Reply that Manolete had put forward a case not based upon the HMRC figures themselves, but using the key constituent elements of those figures to calculate the total sales over the 10 year period of trading.
47. Counsel for Manolete argued that the judge had gone beyond Mr Dalal's evidence and case as presented at trial. I do not accept this. As I have explained, Mr Dalal's case at trial was that the evidence was not sufficiently reliable to permit the court to conclude that there had been underdeclarations of sales by the Company. The judge accepted that case. Counsel for Manolete pointed out that no evidence was adduced by Mr Dalal that the prices in the Watan invoices were unrepresentative of those charged by the Company. It is equally true to say that no evidence was adduced by Manolete that they were representative. The witnesses do not appear to have been asked about this by either side. The Appendix 2 calculation implicitly invited the judge to treat the Watan invoices as representative in the absence of specific evidence either way, and in those circumstances the judge was entitled to accept Mr Dalal's general submission that the evidence was unreliable because of the margins of error.
48. It is nevertheless pertinent to note that, as counsel for Mr Dalal accepted, the lowest prices for which there was documentary evidence by the time of closing submissions were those produced by counsel for Mr Dalal's analysis of the Watan invoices. The evidence from the other five invoices was of higher prices even though they were from an earlier period. (As explained above, the invoices underlying the calculations in the AMS Report were not in evidence, and the information in the Report does not enable averages to be calculated which can be compared with the other figures, but the AMS Report does show that prices rose over time as one would expect.)

Ground 2

49. Ground 2 is a direct challenge to the judge's finding that Manolete had not proved that it is more likely than not that there were additional sales to those declared by the Company. Manolete accepts that this ground faces a high hurdle to overcome for the reasons discussed in many recent authorities such as Volpi v Volpi [2022] EWCA Civ 464, [2022] 4 WLR 48, but it contends that there is an identifiable flaw or gap in the judge's logic and/or an inconsistent treatment of the evidence.
50. In summary, counsel for Manolete submitted that the judge had made findings as to each of the constituent elements in the calculation in Appendix 2. Furthermore, in almost every instance he had accepted the Defendants' case as to the relevant figure. As the judge also found, the result of the calculation was that the total sales were £34,455,432, £3,738,875 higher than those reported, a difference of 12.2%. That difference could not be regarded as de minimis, nor did the judge say that it was. It followed that the judge ought to have found Manolete's case proved to that extent. The only reason the judge gave for not so holding was that the figures were too uncertain, but that reason was belied by the judge's own careful analyses of the evidence. Moreover, the only specific uncertainty the judge identified at [185] was over the sale prices; but not only did the judge himself say that the figures he used were the best evidence available, they were the lowest for which there was any evidence.
51. In my judgment this argument suggests that the judge would have been entitled to find that Manolete had proved its case on the balance of probabilities to the extent of £3,738,875, but it does not demonstrate that he was not entitled to reach the opposite conclusion. My reasons are as follows.
52. First, the judge's starting point, rightly, was Manolete's case as pleaded in its Particulars of Claim. As he noted, that case was squarely based on the HMRC assessments, which in turn were based on method 2 in the BEE. To repeat, that method indicated that there had been total sales of £5,887,092 in the year ending 30 September 2009 compared to £2,695,644 reported in the accounts. The judge therefore reasoned that the first thing he needed to determine was whether the evidence established that there were additional sales in that year. Although the evidence was very thin, he considered there was just enough information for him to make a very rough estimate of the likely level of sales in that year. Having gone through each of the elements of the calculation, he arrived at an estimated figure of around £2.6 million. Thus the exercise did not show additional sales in that year. It followed that Manolete's case as pleaded in its Particulars of Claim failed. Manolete does not challenge that conclusion.
53. Secondly, it is not disputed that it was nevertheless open to Manolete to attempt to prove that there were additional sales in a different way. Thus it is not in dispute that it was open to Manolete to seek to rely upon the calculation in Appendix 2. In considering that calculation, however, account must be taken of the fact that this method of calculation showed no additional sales in 2009. If there were no additional sales in 2009, which was the focus of the HMRC investigation, why should there have been additional sales in other years? The only explanation that was suggested during the course of argument was that the Company fully declared its sales in 2009, but not in the other years. But that would be a paradoxical conclusion for which there is no other evidence. This inevitably undermines the confidence that can be placed in the calculation.
54. Thirdly, although the only specific uncertainty mentioned by the judge at [185] was as to the prices, and although there are reasons for thinking that, in fact, the prices used by the judge were reasonably reliable, the judge did say that this was an example. The more important uncertainties, to my mind, concern the numbers and weights of the chickens. Here Manolete's case raises another paradox. The FSA records as to the numbers and quantities of chickens slaughtered are fundamental to the entire case. The records show that nearly all of the chickens were less than 2 kg in weight, yet most of the other evidence is of chickens weighing more than 2 kg (the exception is the evidence concerning hens and baby birds). At trial no one could explain this discrepancy. The judge adopted HMRC's approach of assuming that the FSA records as to the numbers were correct and taking an average weight of 2 kg. The judge described this as the only reliable basis for carrying out the calculation, but he also described the average weight as a very rough and ready figure. In my view the judge was justified in using these figures in order to test Manolete's case as to the sales in 2009. Once attention moves to the 10-year calculation in Appendix 2, it becomes more important, as the judge said, to consider how soundly-based the figures were for the simple reason that even small errors could be significant once multiplied by ten. The evidence suggests that the FSA records were not accurate as to the weights; but, as the judge observed at [132], if the weights are not accurate it is not possible to have confidence in the quantities either. In any event, a small difference in the average weight could be significant to the overall calculation.
55. Fourthly, if the judge had considered that the Appendix 2 calculation might have some validity, he would have been bound to have weighed it against the evidence of Mr Dalal, Sajid and Anisha. The thrust of Mr Dalal's evidence in his witness statement was to deny that there had been additional sales, and implicitly to deny that there had been unauthorised extractions of cash from the Company. Sajid and Anisha explicitly denied that there were undeclared sales and that they had received any proceeds from such sales. We were told that they maintained their denials in cross-examination. The judge's assessment was that all three were truthful witnesses, albeit that Mr Dalal's evidence was not particularly reliable due to his inability to remember much. It is difficult to suppose, however, that over £3.7 million could have been extracted from the Company without any of the three knowing about it. Certainly, there is no evidence that anyone else could have been to blame.

Ground 3

56. Finally, Manolete contends that the judge impermissibly resorted to the burden of proof, contrary to the guidance summarised in Re A (Children) (Care Proceedings: Burden of Proof) [2018] EWCA Civ 1718, [2018] 4 WLR 117 cited by the judge at [187], to resolve the issue when his own assessment at [117] and the Appendix 2 calculation showed that there was sufficient evidence to enable a rational conclusion to be drawn. I do not accept this contention. As I have explained, there were major gaps in the documentary evidence. The judge did the best he could with the evidence that was available, subjecting it to careful analysis. That analysis enabled him to reach the conclusion that, so far as the available evidence went, there were no additional sales in the year ending 30 September 2009. He also concluded that the evidence did not show that it was more likely than not that there were additional sales in the 10 years of the Company's trading life.
57. Counsel for Manolete relied upon the statement of principle of Toulson LJ in Milton Keynes Borough Council v McNulty [2013] EWCA Civ 15, [2013] 1 WLR 1183 at [35]:

"The civil 'balance of probability' test means no less and no more than that the court must be satisfied on rational and objective grounds that the case for believing that the suggested means of causation occurred is stronger than the case for not so believing."

In my view that is precisely the test that the judge applied.

Conclusion

58. For the reasons given above I would dismiss this appeal.

Lord Justice Newey:

59. I agree.

Lord Justice Moylan:

60. I also agree.
Appendix