Tradition Financial Services Ltd v Bilta (UK) Ltd & Ors

There are three issues raised on this appeal from Marcus Smith J:

The judge’s judgment on the substantive issues is at [2022] EWHC 723 (Ch); [2022] BCC 833. His judgment on costs is at [2022] EWHC 1431 (Ch).

I can take the main underlying facts from the judge’s substantive judgment.

By a claim form issued on 6 July 2015, the following five claimants (by their liquidators) issued proceedings against various defendants:

The liquidators of these companies (Mr Kevin Hellard and Mr David Ingram in respect of Bilta, Weston, Nathanael and Inline, and Mr Kevin Hellard and Mr Ian Richardson in respect of Vehement) were also named as claimants.

The defendants, at that stage, comprised (i) SVS Securities plc (“SVS”), (ii) a former director of SVS (Mr Virk); (iii) a former employee of SVS (Mr Fox); and (iv) Deutsche Bank AG (“Deutsche Bank”).

The claims against SVS, Mr Virk, Mr Fox and Deutsche Bank all concerned missing trader intra-community fraud (“MTIC Fraud”) which took place in the summer of 2009.

MTIC Fraud exploits the fact that imports from one EU country into another are VAT-free. The most basic form of MTIC Fraud involves traders importing goods VAT-free from elsewhere in the EU, selling them within an EU country with VAT added to the sale price and so running up large liabilities to account for the VAT to national revenue authorities. These importers then default on their liabilities to account for VAT, instead paying their VAT receipts away to third parties and going into insolvent liquidation.

MTIC Fraud thus concerns a failure, on the part of the responsible company, to account for its output VAT, which is the VAT that a company must calculate and collect when it supplies goods or services. Output VAT is calculated both on supplies to other businesses and supplies to ordinary consumers. By contrast, input VAT is the VAT added to the price when goods or services that are liable to VAT are purchased. If the person or business that is buying the goods or services in question is registered for VAT, then the input VAT may be claimed back from HMRC. Generally speaking, businesses registered for VAT will incur both input and output VAT, and the net amount will be paid to HMRC (in the case of the United Kingdom).

Those involved in MTIC Fraud will usually seek to transact in high-value, easily transportable products. Mobile phones, SIM cards and computer chips were popular product vehicles for MTIC Fraud in the early- to mid-2000s. The fraudsters involved tend to transact in high volume and at high speed and frequency, using back-to-back transactions between linked companies set up or acquired for the purpose. The transaction chains may also include companies that are not controlled by the fraudsters. The use of a chain of companies obscures the fraud and complicates the task of investigation.

In these proceedings, the MTIC Fraud involved spot trading in carbon credits under the EU Emissions Trading Scheme (“EUAs”). One EUA represents the right to emit one metric tonne of carbon dioxide or other specified gas. EUAs are required by so-called “installations” in energy intensive industries that emit carbon dioxide and are subject to obligations to comply with EU emissions rules. EUAs exist only in electronic form, are fungible, and can be traded almost instantly. Spot trading in EUAs within EU Member States attracted VAT in 2009. Those rules changed when the authorities realised that EUAs were being used as the products in MTIC Fraud.

As a result of the MTIC Fraud, the claimants were left with enormous VAT liabilities owing to HMRC. In the case of all of the claimants, HMRC were and are the primary creditor.

The ordinary limitation period in respect of these claims ended by the summer of 2015. The claims against SVS, Mr Virk, Mr Fox and Deutsche Bank were all issued just within the ordinary limitation period. We are no longer concerned with those claims.

On 8 November 2017 the claimants issued a separate claim form against Tradition Financial Services Ltd (“TFS”). The claims against TFS were: (i) claims by the companies themselves alleging TFS’s dishonest assistance in the breach of fiduciary duty by the directors of the claimants; and (ii) claims by the liquidators alleging participation by TFS in the fraudulent trading of the businesses of the claimant companies pursuant to section 213 of the Insolvency Act 1986.

The dishonest assistance claims alleged that through spot EUA trading carried out between May and July 2009, TFS participated in chains of transactions by which a large MTIC Fraud was perpetrated. That participation amounted to dishonest assistance by TFS in the breaches of fiduciary duty of the directors of the claimants, which those directors owed to those companies.

As a consequence of a partial settlement agreement between the claimants and TFS, there were two issues for the judge to decide:

On this first of those questions, the judge held that the claims against TFS were indeed statute barred. Although the claimants had relied on section 32 of the Limitation Act 1980 (which postpones the running of time where, in the case of an action based on the defendant’s fraud, the claimant could not have discovered the fraud with reasonable diligence), the judge found against all of the claimants on the facts. On the second of those questions, the judge held that TFS were within the scope of section 213. The claims brought by the liquidators of Nathanael and Inline therefore succeeded under that section; but the claims by the liquidators of the other claimants did not because they were statute barred. Accordingly, the claims brought by Bilta, Weston and Vehement (and their liquidators) failed completely; and there is no appeal in relation to their claims. The claimants appeal against the first of the judge’s conclusions in relation to Nathanael and Inline; and TFS cross-appeals against the second. We heard argument on these two issues in reverse order.

The costs appeal is, to a large extent, contingent on the outcome of the two substantive issues.

Claims under section 213 of the Insolvency Act 1986 were brought by the liquidators of all five companies.

The judge did not make (and was not asked to make) any findings of fact about TFS’s precise role. The question that was placed before him under the terms of a consent order of 21 February 2022 was whether “on the facts pleaded” the claimants were entitled to the declarations claimed under section 213. The “facts pleaded” referred to the facts pleaded by the claimants, even though many of those alleged facts were disputed by TFS. Since the agreed issue referred to “the facts pleaded” (by the claimants), before embarking on a consideration of the law, I should set out the assumed facts.

In brief the assumed facts were as follows. The five claimant companies all became insolvent owing sums to HMRC in respect of VAT. They were involved in the MTIC Fraud described by the judge, either as defaulters (Nathanael and Inline) or because they took part in deal chains that led to defaulters (i.e. chains that led to Bilta, Weston or Vehement). SVS bought EUAs on a back-to-back basis for those companies.

TFS brokered deals on behalf of companies (i.e. its clients) that dealt with SVS on a back-to-back basis. TFS’s role in brokering a deal was to find a counterparty and negotiate the terms on which that counterparty was prepared to sell EUAs to or buy EUAs from the client company via SVS. TFS was paid brokerage by volume traded.

TFS participated in sales by SVS in the following ways. First, in the case of some of the deals to which the claim related, TFS introduced SVS to its overseas suppliers and liaised between SVS and the overseas supplier for the purpose of informing buyer and seller of the price and amount in which each was willing to trade. Second, it introduced Nathanael and Inline to SVS. That was a fact relevant to only some of the EUA trades to which the claim related. Third, for those clients introduced to SVS, TFS liaised between those companies and SVS for the purpose of informing SVS of the price at which they were willing to sell and the amount that they wished to sell. Overall, TFS were involved in a number of the deal chains, but not all of them.

TFS introduced Inline and Nathanael to SVS, knowing that they were unlikely to be legitimate trading concerns, without any belief that they would be engaging in spot trading of EUAs for legitimate purposes and in the knowledge that their purpose was to amass VAT.

TFS offered the suppliers of the EUAs who wanted to sell with VAT the prospect of selling through SVS with an expectation of payment the same day thereby enabling them to buy and sell further EUAs with the monies received. Payment on the same day or even the next day was unusual and limited the number of counterparties to whom SVS could sell, thus depressing the price that could be obtained for the EUAs.

TFS was aware of the likelihood that companies newly applying to trade EUAs from around May 2009 were fronts for individuals wishing to conceal their involvement (or the extent of their involvement) in spot trading of EUAs and the likelihood that there was no legitimate reason for their wishing to do so.

TFS did not perform any genuine “know your client” (“KYC”) inquiries of the suppliers aimed at reducing the possibility that the suppliers were involved in any criminal activity before asking SVS to provide a conduit for their EUAs to exporters. TFS’s traders knew this, and they also knew that SVS did not perform any genuine KYC inquiries.

TFS was aware that the nature and pattern of the EUA trading in which SVS and TFS were involved was suspicious and such as to call for inquiry as to whether the trading was legitimate; and knew, or did not care whether, such trading was linked to a financial crime (including money laundering) and in particular a VAT fraud. Despite its suspicions, TFS failed to make inquiries with a view to trying to ascertain whether its clients’ trading was legitimate and not connected with criminal activity in particular VAT fraud and whether or not to trade (or continue to trade) with such companies if there was reason to think it was.

To the contrary, TFS pretended it had credible explanations for its clients’ trading though neither of the explanations relied on were honestly thought to be adequate explanations for the trading.

Bilta was ordered to be wound up on 25 November 2009, and the liquidators were appointed on the same day. Nathanael was ordered to be wound up on 19 March 2012 and the liquidators were appointed on 19 August 2013. Weston was ordered to be wound up on 27 October 2010 and the liquidators were appointed on 13 June 2011. Vehement entered creditors’ voluntary liquidation on 23 October 2009 and the liquidators were appointed on 18 March 2010. Inline was ordered to be wound up on 8 June 2015 and the liquidators were appointed on the same day.

Section 213 of the Insolvency Act 1986 provides:

Because the application can only be made in the course of a winding up, and only by the liquidator, it is common ground that the cause of action arises when a winding up order is made, or the company goes into voluntary liquidation. Once the cause of action has arisen, the liquidator then has six years to bring the claim: Limitation Act 1980 s 9; Re Overnight Ltd [2009] EWHC 601 (Ch), [2009] Bus LR 1141. Since the claim form was issued on 8 November 2017, it follows that in the case of any company where the winding up order was made before 8 November 2011, any cause of action under section 213 is statute barred. That applies to the claims made by the liquidators of Bilta, Weston and Vehement.

The issue on this appeal is the scope of the words “any persons who were knowingly parties to the carrying on of the business in the manner above-mentioned”. TFS argued that that phrase was restricted to persons exercising management or control over the company in question. The judge rejected that argument, which TFS resurrects on this appeal.

In the course of his oral submissions Mr Scorey KC, for TFS, did suggest that even if section 213 was not so limited, nevertheless TFS, on the facts of this case, was not liable under that section. That would involve a detailed evaluation of the facts relating to TFS, which the judge did not carry out; and which was not foreshadowed in the grounds of appeal or the skeleton argument. In my judgment Mr Parker KC was justified in objecting that this appeal is concerned only with the narrow question of law I have identified.

The argument for TFS rests, in part, on a detailed examination of the legislative history of provisions relating to fraudulent trading by a company in liquidation; and a closely reasoned article by Mr David Foxton QC: Accessory liability and section 213 Insolvency Act 1986 [2018] JBL 324.

There have been provisions enabling the court to set aside transactions effected with intent to defraud creditors as far back as the Fraudulent Conveyances Act 1571. But the statutory history of corporate fraudulent trading begins with section 75 of the Companies Act 1928. That section applied if, in the course of a winding up, it appeared that any business of the company had been carried on with intent to defraud creditors or for any other fraudulent purpose. Where that was established, the court was empowered to declare “any of the directors, whether past or present, who were knowingly parties to the carrying on of the business as aforesaid” to be personally responsible for the debts of the company. Section 75 (5) in effect defined “director” to include what we would now call a shadow director. Section 75 (3) imposed a criminal sanction on a director within section 75 (1). TFS argues that this was, in effect, a “veil-piercing provision”, limited to directors as defined, and was not intended to (and did not) create a secondary form of accessory liability on third parties. Section 75 was re-enacted, with minor and irrelevant changes, in section 275 of the Companies Act 1929.

In 1943 the President of the Board of Trade appointed a committee under the chairmanship of Cohen J to consider what major amendments were desirable in the Companies Act 1929. The Cohen Committee reported in 1945 (Cmd 6659). Their recommendation in relation to section 275 was:

Parliament accepted that recommendation; and made the legislative changes by section 101 of the Companies Act 1947, although not in precisely the same terms as the recommendation. Shortly afterwards, the companies legislation was consolidated in the Companies Act 1948. Section 332 of that Act provided:

TFS argues, supported by Mr Foxton’s article, that the object of the Cohen Committee’s recommendation was to deal with a case where the company’s business was directed, controlled or managed by a criminal mastermind who was the brains behind the fraud and where the directors were merely dupes. Mr Foxton quotes some of the evidence given to the Committee on that topic. Although it shows that the trigger for the thought process that led to the recommendation was indeed as TFS suggests, the inquiry itself went further. Thus, Mr Foxton quotes Sir Lionel as asking two distinguished witnesses whether it would be useful for the section to apply “to any other party who was knowingly a party to carrying out the fraud”. Moreover, the Committee’s ultimate recommendation was not limited to the criminal mastermind behind the scenes. We have no way of knowing what deliberations of the Committee took place in private before it made its recommendation. It is also to be noted that the recommendation extended to persons who were “knowingly parties to the fraud” not merely those who were involved in the running of the fraudulent business. Since the legislation defined “director” in terms that included what we now call a “shadow director”, it must have been the Committee’s intention to widen the net beyond those in active control of the business (whether formally directors or not).

The next relevant iteration of the legislation was the Companies Act 1985. The principal change effected by that Act (apart from a more modern style of drafting) was to separate the criminal liability and the civil liability into two sections (sections 458 and 630 respectively). In addition, under section 458 the criminal offence could be committed even if the company was not being wound up; thus departing from what the House of Lords had held in DPP v Schildkamp [1971] AC 1 (this amendment had in fact been made by the Companies Act 1981). In that respect, therefore, criminal liability was more extensive than civil liability.

Further amendments were made by Schedule 6 to the Insolvency Act 1985. Paragraph 6 of that Schedule inserted a new section 630 (2):

This amendment thus did two things. First, it narrowed the class of persons who could apply under the section. Whereas previously a creditor or contributory was able to apply, following the amendment the liquidator was the only person empowered to do so. Second, in place of the defendants being personally responsible for the company’s debts and liabilities, the court now had power to declare them liable to “make contributions” to the company’s assets.

The provisions relating to civil liability are now to be found in section 213 of the Insolvency Act 1986 (quoted above) and those relating to criminal liability are in section 993 of the Companies Act 2006. Section 993 (1) provides:

Conviction of the offence may result in a prison sentence of up to 10 years or a fine (or both). Section 993 (2) provides that the criminal offence may be committed whether or not the company has been or is being wound up. Section 213, on the other hand, only applies to a winding up.

To bring the legislative history up to date, section 9 of the Fraud Act 2006 created a parallel offence in relation to non-corporate businesses. The explanatory notes to that section state that case law establishes that the offence:

Provisions equivalent to section 213, but relating to companies in the course of administration, are now to be found in section 246ZA of the Insolvency Act 1986 (introduced by the Small Business, Enterprise and Employment Act 2015, section 117).

In Re Maidstone Buildings Provisions Ltd [1971] 1 WLR 1085 the company went into insolvent liquidation. Mr Penney was its company secretary. The allegation against him was that as well as being the company secretary, he was also its financial adviser, was present at board meetings and took no steps to prevent the company from continuing to trade. Pennycuick V-C held that Mr Penney was not liable under section 332 of the 1948 Act. He said at 1092:

He added at 1093:

This decision does give some slender support to a narrow interpretation. First, it seems to me that one of the questions for the Vice Chancellor was whether Mr Penney was “concerned in the management in the company”. Second, the Vice Chancellor considered that the argument that the section embraced everyone with notice of the fraud would lead to “impossible results”. On the other hand, all that the Vice Chancellor actually decided was that active participation was necessary; and that concurrence in the carrying on of the fraudulent business might be enough. By that he meant “taking some positive steps in the carrying on of the company's business in a fraudulent manner”. He did not say that those positive steps must amount to management or control.

In Re Murray-Watson Ltd (unreported) 6 April 1977 Oliver J said that section 332:

This observation takes a relatively narrow view of what amounts to carrying on a fraudulent business, but it does not, I think, shed much light on who may qualify as being party to the carrying on of that business. Unfortunately, it has not been possible to find a transcript of that case, and all that we have to go on is the quoted extract from the judgment.

In Re Gerald Cooper Chemicals Ltd [1978] 1 Ch 262 the respondents were a company and its directors who had lent money to the insolvent company. The loan had been repaid out of money raised by purported prepayments for the sale of goods under orders which could not be fulfilled, at a time when the company was already insolvent. Templeman J held that the company was carrying on business with intent to defraud creditors by accepting deposits knowing that they could not supply the goods and were insolvent. He went on to say at 268:

Although the Maidstone case was cited to Templeman J he did not mention it; and he seems to me to have taken a broader view of the scope of section 332 than the Vice-Chancellor.

Mr Scorey placed some reliance on the decision of the Court of Appeal, Criminal Division in R v Grantham [1984] 1 QB 675. Mr Grantham registered a company as carrying on the business of importing root crops. He was the sole authorised signatory of the cheques and was the “consultant” in charge of administration and holder of the purse strings. The company traded at a loss. The trial judge directed the jury:

The trial judge’s direction did not require the prosecution to prove that Mr Grantham exercised management or control of the business: merely that he took an active part in carrying it on. The jury decided all three questions in favour of the prosecution. Mr Grantham appealed, but he only appealed on the question whether he had the necessary intent to defraud. He did not appeal on the question whether he had taken an active part in the business, even though his case on the facts was that he had not. For that reason, the Court of Appeal did not consider the trial judge’s formulation of the first question. It is fair to say, however, that in the course of his judgment Lord Lane CJ said at 683:

I do not read the reference to “whoever was running the business” as casting any real light on the limits of the offence. In those circumstances it does not seem to me to be an authority of any real weight on that question.

Re Augustus Barnett & Son Ltd (1986) 2 BCC 98,904 concerned an attempt to impose liability under section 332 on the parent company of an insolvent company. The parent had subsidised its subsidiary; and had also announced that it would continue to support it financially. But when it withdrew its support the subsidiary went into insolvent liquidation. Citing Re Gerald Cooper Chemicals Ltd, Hoffmann J said at 98907 that the section required:

Hoffmann J seems to me to have analysed the section as imposing a primary liability on those carrying on a fraudulent business, and an accessory liability on persons who in some way participated in the fraudulent acts. That analysis is consistent with the recommendation of the Cohen Committee, and contradicts TFS’s argument.

But he went on to say that since no fraudulent intent was alleged against any person who actually carried on the business, there were “no fraudulent acts to which the outsider can have been a party and his own state of mind seems to me for present purposes irrelevant.” He added:

Clearly Hoffmann J envisaged that a person could be liable as an accessory under section 332 even though not actually running the business.

By the time that R v Miles [1992] Crim LR 657 came to be decided the criminal offence had been hived off into section 458 of the Companies Act 1985. Mr Miles was charged with an offence under section 458. The Crown’s case was that he had been selling worthless shares on behalf of a company. The Crown also asserted that Mr Miles had a managerial role in the company, whereas he contended that he was merely a salesman acting under orders. The trial judge directed the jury that they had to be sure that Mr Miles was “participating, or taking part or concurring in the actual trade which is involved in the business of the particular company”. The Court of Appeal, Criminal Division held that that was a misdirection. But it is important to understand the submissions made by Mr Barnes QC on behalf of Mr Miles in order to evaluate the court’s decision. Watkins LJ summarised them as follows:

Giving the judgment of the court Watkins LJ said:

The court referred to the Maidstone case in the course of the judgment but not to Gerald Cooper. What is important, however, is that the proposition that the court accepted was not Mr Barnes’ first proposition (i.e. some controlling or managerial function is required) but his second (i.e. if there is an issue about whether a defendant who is working for the company comes within the section clear, simple guidance should be given). Had Mr Barnes’ first proposition been accepted, it is difficult to see what further clear and simple guidance would have been needed. Moreover, the particular part of the summing up to which the court took exception, particularly in the way that the prosecution had put its case, was the trial judge’s direction that it was enough that Mr Miles “concurred” in the trade. The court did not criticise the trial judge’s direction that the jury had to be sure that the defendant was “participating, or taking part” in the carrying on of the business. One can concur in a trade without actively participating in it. To that extent, therefore, the Court of Appeal restricted the meaning that the Vice-Chancellor had adopted in the Maidstone case. I consider that we can accept that in the context of civil liability as well as criminal liability, concurrence is not enough. Although it is true that the court referred to Lord Lane’s dictum in Grantham, they did not say that the offence was limited to those running the business; merely that it included such persons.

In R v Kellard [1995] 2 Cr App 134 a number of defendants, including Mr Wright, were charged with fraudulent trading. The relevant counts, for present purposes, were counts 14 and 15. Counsel for Mr Wright submitted that the Crown’s case on those counts was very weak because Mr Wright was neither a director nor employee of the companies concerned and could not be regarded as exercising a controlling or managing function or as one who was running the business (referring to Miles). Farquharson LJ said in relation to those counts:

The reason why Mr Wright was guilty was because he was acting on behalf of the company in question. The test was formulated, not as running or controlling the business, but “assisting” in carrying it on.

There the matter stood when Parliament enacted section 213. I digress at this point because Mr Parker KC relies on the Barras principle in support of his argument in favour of the wider meaning. The Barras principle derives from Barras v Aberdeen Steam Trawling and Fishing Co Ltd [1933] AC 402, 411 in which Viscount Buckmaster said:

Mr Parker submits that the interpretation given to section 332 in Gerald Cooper is the relevant clear judicial interpretation and the interpretation that Parliament must be taken to have adopted. I do not agree. In my judgment, the diversity of views expressed in the cases do not engage the Barras principle. I do not consider that, taken together, they can be said to amount to a “clear” judicial interpretation. Moreover, Gerald Cooper was a case decided by a first instance judge on a strike out application; and it cannot be assumed that Parliament had it in mind when re-enacting section 332 (see the example given by Lord Blanesborough, recounting his experience as a pupil, in Barras at 433-434). So, I do not feel able to place any reliance on the Barras principle. In addition, in Morphitis v Bernasconi [2003] EWCA Civ 289, [2003] Ch 552 (decided after section 213 came into force) Chadwick LJ commented at [46] that Gerald Cooper may have adopted too expansive a view of section 332. Nor do I consider that, as Mr Parker submitted, the limited amendments made by the Insolvency Act 1985 can be said to represent an endorsement of Gerald Cooper. The most that can be said is that Parliament did not explicitly repudiate the wide construction adopted in Gerald Cooper. But given the decision in Miles, Parliament might have taken the view that there was no need to do so.

The next case chronologically, and a very influential one, is the decision of Neuberger J in Re BCCI [2002] BCC 407, although it has been subjected to considerable criticism by Mr Foxton in his article. The liquidators of BCCI alleged that another bank (BAII) had (1) knowingly participated in BCCI’s secret and illegal acquisitions and/or financing of shares in an American bank, (2) knowingly participated in BCCI’s secret and illegal acquisition and/or financing of the acquisition of an American bank, and (3) knowingly participated in BCCI's fraudulent overstatement of assets and earnings by assisting BCCI artificially to service certain non-performing loans in its books thereby avoiding having to make any provision for the loan. The procedural question was whether Neuberger J should determine a preliminary issue which he reformulated as follows. BAII’s contention was that:

Neuberger J held:

It is to be noted that what Neuberger J actually decided was not that the claim against BAII would succeed. He merely decided that it was not bound to fail. His decision therefore leaves open the precise limits of the scope of section 213.

Neuberger J gave six reasons for his conclusion. His first reason was that:

Mr Scorey takes issue with this reason. He argues that to be party to the carrying on of a business, a person must take part in its carrying on. That, he says, is how the word “party” is used in other legal contexts. A party to a contract, for instance, is someone with whom the contract is made, and who has legal rights and obligations under it. A party to legal proceedings is a person who has formally been joined in those proceeding. A party to a marriage is one of the persons who are married to each other.

Mr Parker disputes this. He argues that as a matter of ordinary English a person who knows that the business he is dealing with is fraudulent is a party to the fraudulent carrying on of that business. The relevant sense of the phrase “party to” is that given by the Oxford English Dictionary (online edition):

In my judgment, the phrase is capable of bearing either meaning. But even on Mr Scorey’s argument it does not, in my judgment, follow that in order to take part in the carrying on of a business one must have a controlling or managerial function in it.

Neuberger J’s second reason was based on policy. Although he said that it would be wrong to interpret section 213 as so wide as to stultify “normal business transactions”:

Mr Scorey submits that policy-based reasoning is not appropriate. He accepts that everybody agrees that fraud is “a bad thing” but what constitutes fraud, and the legal remedies available, are questions that should be resolved by legal principles rather than policy.

It is no doubt true that questions of interpretation of statutes should not rest wholly on questions of policy, rather than the statutory language, but policy does have a role to play. In Bank of India v Morris [2005] EWCA Civ 693, [2005] BCC 739, for example, Mummery LJ, giving the judgment of this court said at [93]:

The court went on to discuss the policy underlying section 213 at [108] to [112].

Neuberger J’s third reason was the contrast in wording between section 213 on the one hand, and sections 212 and 214 on the other. Section 212, however, does not create any liability where none existed before. As Sealy & Milman put it in The Annotated Guide to the Insolvency Legislation 2022:

As far as section 214 is concerned, the impulse behind its enactment was the difficulty of establishing dishonesty under the predecessors of section 213. Nevertheless, the difference in wording is, in my judgment, a legitimate factor to take into account.

Neuberger J’s fourth reason was the legislative history of section 213. He referred in particular to the recommendation of the Cohen Committee that the section be extended to “other persons who were party to the fraud”. In his article, Mr Foxton is critical of this reason. He points out that Neuberger J did not refer to the evidence which had prompted the Committee’s recommendation. That is true, but for the reasons I have explained, I do not consider that the criticism is a weighty one.

Neuberger J’s fifth reason was that it was consistent with first instance authority; and his sixth reason was that the decisions of the Court of Appeal, Criminal Division, were not directed to the particular point that arose in BCCI. It is convenient to take these reasons together. In his article, Mr Foxton is particularly critical of Neuberger J’s treatment of the latter cases. Dealing with the criminal cases, Neuberger J said of them:

For reasons I have explained, I do not consider that that is what they (and especially Miles) decided. Neuberger J went on to say:

Neuberger J went on to say:

I think, with respect, that this does not adequately deal with Miles. What, in my judgment, the court held in Miles was:

I do not think that the case turned on the fact that Mr Miles was only an employee and for that reason could not be guilty of the offence. In my judgment Neuberger J took an unnecessarily narrow view of what Miles decided.

Mr Foxton writes of this part of Neuberger J’s judgment:

As will be appreciated, I do not agree with Mr Foxton’s analysis of Miles either. Mr Foxton in effect says that the court in Miles accepted Mr Barnes’ first proposition. It did not. It accepted his second proposition.

Mr Scorey also points to what he says are the odd results of Neuberger J’s reasoning. If those managing a company carry on the business in a fraudulent manner and instruct an employee to complete transactions in order to further that purpose, then even if the employee knows about the fraudulent purpose, he will not be liable. On the other hand, a counterparty to those transactions will be liable even if he has no more than blind-eye knowledge. This submission depends on reading across the decision in Miles as regards the criminal offence into the scope of civil liability. It also presupposes that the court in Miles accepted Mr Barnes’ first proposition, which in my judgment it did not. But in any event Miles does not occupy the whole of the field. Quite apart from the offences to which Mr Foxton referred, the company’s employee who knows about the fraudulent intent and furthers that purpose could be charged with aiding and abetting fraudulent trading: R v Cartwright [2020] EWCA Crim 369, [2020] Lloyd’s Rep FC 547. In addition, Neuberger J did not purport to define the scope of section 213. He merely said that the claim in that case was not bound to fail.

As Mr Scorey also points out, Neuberger J gave another reason for his decision (albeit obiter). He said:

Thus, the actual result of the case was that Neuberger J dismissed the application for the trial of a preliminary issue, with the consequence that the action was permitted to go forward to trial.

Morphitis v Bernasconi was concerned with what amounts to “the carrying on of a business with intent to defraud”, rather than the question who would be “party” to the carrying on of such a business. Chadwick LJ said at [46]:

But this case, together with the decision of Oliver J in Re Murray-Watson Ltd does support the proposition that in order to liable under section 213 the defendant must have been party to the carrying on of a fraudulent business; and not merely a party to fraud. I agree with that. What must be established is that the company’s business (or a discrete part of that business), taken as a whole, had been carried on fraudulently. Whether that is established in any particular case is a question of fact. Mr Scorey submitted that if the defendant was carrying on his own business, he could not be said to be party to the carrying on of a different business. I do not agree. The extent to which a counterparty must be involved in the carrying on of the fraudulent business may depend upon the facts. Suppose that a manufacturer regularly supplies counterfeit designer clothes to a retailing company, knowing that the retailer will pass them off as genuine. It is, in my judgment, no misuse of language to describe the manufacturer as “party to the carrying on” of a fraudulent business, even though he exercises no managerial or controlling role within the retailing company; and the manufacturer may have other business activities that are not fraudulent. The manufacturer knows about the retailer’s fraudulent business and is actively participating in it in the sense of furthering and facilitating it. In my judgment, the allegedly odd results are more apparent than real.

Bank of India v Morris (to which I have already referred) was another case arising out of the collapse of BCCI. BCCI’s liquidators brought a claim against the bank alleging that they were knowingly party to the carrying on by BCCI of a fraudulent business. The real issue in the case was whether the knowledge of a senior manager of the London branch of the Bank of India (referred to as “SBI”) could be attributed to that bank. Patten J and this court both answered that question “yes”. At first instance ([2004] EWHC 1868 (Ch), [2004] 2 BCLC 236) Patten J said at [12]:

In this court, ([2005] EWCA Civ 693, [2005] BCC 739) as Mummery LJ recorded at [86], the bank accepted that section 213 could apply to a corporate counterparty. This point reappeared at other parts of the judgment:

The court considered the fact that since 1985 criminal liability and civil liability had been split into two different sections. As to that, they said at [107] that the separation of criminal and civil liability made it easier to isolate and focus on the policy of imposing civil liability. The court discussed the policy underlying section 213, which was the subject of argument. At [111] the court said:

They added at [120]:

At [129] the court returned to the inter-relationship between civil and criminal liability. They said:

The next case to which I should refer is the decision of the Supreme Court in Bilta (UK) Ltd v Nazir (No 2) [2015] UKSC 23, [2016] AC 1. The relevant defendants in that case were Jetivia SA (a Swiss company) and a Mr Brunschweiler, its sole director and a resident of France. It was said that they were liable both in conspiracy and dishonest assistance in a breach of fiduciary duty and also under section 213. The allegations against those defendants were not entirely easy to discern from the judgments in the Supreme Court in so far as concerns section 213. But they are set out in more detail in the judgment at first instance of Sir Andrew Morritt V-C: [2012] EWHC 2163 (Ch), [2014] Ch 52 at [7]. The overall case, like the present one, concerned an MTIC fraud in relation to trading in EUAs. Bilta bought EUAs from Jetivia, free of VAT, and sold them in the UK with VAT at prices net of VAT that were lower than the price they had paid. It was also alleged that the pattern of trading between Jetivia and Bilta was not bona fide or consistent with legitimate commercial trading and that Jetivia knew that its dealings with Bilta were dishonest and part of an MTIC fraud. The particular point raised before the Supreme Court was whether section 213 had extra-territorial effect. The Supreme Court unanimously held that it did. Most of the case concerned the scope of a defence of illegality. But in their joint judgment Lords Toulson and Hodge introduced section 213 as follows at [213]:

They added at [215]:

Although the point does not appear to have been argued, it seems to me that Jetivia’s connection with the MTIC fraud was no more intimate than that of TFS in the present case.

In Bilta (UK) Ltd v NatWest Markets Ltd [2020] EWHC 546 (Ch) Snowden J said at [186] that the statements of principle in Bank of India v Morris were “clear Court of Appeal authority” for the proposition that section 213 is not limited to those who have been involved in the management of the company whose business has been carried on with intent to defraud, but potentially extends to outsiders who deal with the company.

Mr Parker submits that Snowden J was right, and that we are likewise bound by those statements of principle. The judge below took a more nuanced view. He noted at [76] that the choice between the narrow and the wide construction of section 213 was not argued in this court in Bank of India v Morris. He regarded that case as no more than persuasive authority.

In my judgment the judge was correct; and we are not strictly bound by the observations of this court in Bank of India v Morris. Whether a subsequent court is bound by a decision of a previous court on a point that was assumed to be correct, but which was not argued, is the subject of comprehensive discussion in R (Kadhim) v Brent London Borough Council Housing Benefit Review Board [2001] QB 955. At [21] and [22] Buxton LJ explained that there were several ways in which a proposition of law could become part of the ratio of a case in the sense that it was a necessary step in the court’s reasoning. One such way was by agreement or concession. The important point to note is that Buxton LJ’s subsequent discussion of the authorities is in the context that the proposition of law in question is part of the ratio of the earlier case. After a review of the authorities, Buxton LJ said at [33]:

He then proceeded to explain the justification for that conclusion. In R (Youngsam) v Parole Board [2019] EWCA Civ 229, [2020] QB 387, to which we were referred, Kadhim was cited with approval but only for the method of identifying the ratio of a case. I do not read Youngsam as casting any doubt on the proposition stated in Kadhim at [33]. Moreover, in the later case of FSHC Group Holdings Ltd v GLAS Trust Corpn Ltd [2019] EWCA Civ 1361, [2020] Ch 365 (referring to Kadhim among other cases) this court held at [136]:

In my judgment, therefore, there is no authority binding on this court that holds that the wider interpretation of section 213 is correct. As I have said, the judge correctly said that there was no precedent binding on him, but that he should follow decisions of courts of co-ordinate jurisdiction unless persuaded that they were clearly wrong. That is an entirely conventional approach to authority: Colchester Estates (Cardiff) Ltd v Carlton Industries plc [1986] Ch 80; Re Cromptons Leisure Machines Ltd [2006] EWHC 3583 (Ch), [2007] BCC 214. The judge said at [79]:

The modern approach to statutory construction is to have regard to the purpose of a particular provision and interpret its language, so far as possible, in a way which best gives effect to that purpose: Barclays Mercantile Business Finance Ltd v Mawson [2004] UKHL 51, [2005] 1 AC 684 at [28]. The purpose of section 213 is the compensation of those who have suffered loss as a result of fraudulent trading. It is to make those who have been parties to fraudulent trading liable to compensate the creditors of the fraudulent company: Bank of India v Morris at [111] and [120].

Mr Scorey rightly submits that where contravention of a statutory provision entails both civil and criminal liability, the court should strive to avoid an interpretation which penalises someone where the legislator’s intention to do so is doubtful. The difficulty with that submission is this court’s conclusion in Bank of India v Morris that the statutory severance of criminal and civil liability has had the effect that a decision on the scope of civil liability has no effect on the scope of criminal liability. On the basis that that conclusion is correct, that leads to the further conclusion that, even if my reading of them is wrong, cases such as Miles and Kellard do not cut down the scope of civil liability under section 213. Although it might be anomalous for criminal liability to be wider than civil liability, it is not necessarily anomalous for civil liability to be wider than criminal liability, particularly where the statutory provisions are now contained in different sections and different Acts of Parliament. Moreover, as noted, the criminal offence may be committed even if the company is not in the course of winding up. Quite apart from that, a person with no managerial or controlling role within a company can be convicted of aiding and abetting fraudulent trading. If there is no objection to a person committing a criminal offence by aiding and abetting fraudulent trading, it is hard to see why there should be any objection to such a person incurring civil liability for the same actions.

I am prepared to accept that, as Mr Scorey submitted, at the statutory birth of this form of liability it was seen as principally concerned with removing the protection of corporate personality. It was, as he put it, a form of “veil-piercing” that made company directors liable for the company’s own debts. But I consider that the deliberate widening of the scope of the section following the recommendation of the Cohen Committee, and the substitution of a liability to make contribution to the company’s assets in place of a liability for the company’s own debts have greatly weakened that link. The purpose of section 213, as it has evolved, is not limited to veil-piercing. It is to secure compensation for those who have suffered loss as a result of the fraudulent trading. As Dixon CJ explained in the High Court of Australia in Hardie v Hanson (1960) 105 CLR 451, 459 (cited by Mr Foxton), the original section:

That explanation has become even less tenable once the Insolvency Act 1985 substituted a liability to make contributions in place of personal responsibility for the company’s debts.

Mr Scorey referred to an article by Mr Patrick Ussher on the Irish equivalent of section 213: Fraudulent Trading (1984) 6 Dublin ULJ 58. He begins by saying that the Irish legislation is “a means by which the cloak of limited liability may be stripped away”. But as the author goes on to say:

I accept also that, as Mr Scorey submitted, it is a precondition of liability that business has been carried on with fraudulent intent; and that incidental frauds committed in the course of a business are not enough. But is it clear in this case that the claimant companies who orchestrated the MTIC fraud were carrying on business with fraudulent intent. The question is who else can be made liable. As I have noted, the recommendation of the Cohen Committee was that liability should be extended to those who participated “in the fraud.” In Mr Ussher’s words, liability extends to those who were party to “the impugned acts” (i.e. the fraud). That does indeed appear to be the position in Ireland. In O’Keeffe v Ferris [1997] 3 IR 463 (cited by Mr Foxton) O’Flaherty J (giving the judgment of the Irish Supreme Court) said at 469:

This observation also distinguishes between primary liability and accessory liability. That is the same distinction that, in my view, Hoffmann J drew in Augustus Barnett. It is also pertinent to note that the observations of Patten J in Morris v State Bank of India about secondary (or accessory) liability (quoted at [94] above) were expressly approved by the Court of Appeal (Criminal Division) in R v Hunter [2021] EWCA Crim 1785, [2022] 3 WLR 485 at [128].

As I have already said, the phrase “party to” can bear either of the two meanings for which Mr Scorey and Mr Parker respectively contended. It is, in my judgment, more consonant with the purpose of section 213 to interpret that phrase in the wider rather than the narrower sense. It is also more consistent with the line of authority to which I have referred. Any gap between the civil law and the criminal law is filled, in the case of the criminal law, by the offence of aiding and abetting.

Mr Scorey emphasised a number of potentially unsatisfactory consequences of adopting the wide interpretation of section 213. First, there is the question of limitation of actions. Under ordinary circumstances, a dishonest assistant in a breach of fiduciary duty is entitled to the benefit of a six year limitation period, unlike a true fiduciary: Williams v Central Bank of Nigeria [2014] UKSC 10, [2014] AC 1189. There are good policy reasons, he said, why true fiduciaries should be liable for fraudulent breaches of trust without limit of time, which do not apply to dishonest assistants. By analogy, there may be good reason why time should not begin to run under section 213 until a company is in the course of winding up as against those who were in control of the company’s fraudulent business, but those reasons do not apply to those who were no more than dishonest assistants. Even a fraudster is prima facie entitled to the benefit of a limitation period.

Second, there is the question of insolvency set-off under rule 14.25 of the Insolvency (England and Wales) Rules 2016. If a dishonest assistant is sued in equity, Mr Scorey says that he will be entitled to set off any debt due to him by the company incurred before the onset of liquidation. By contrast, if sued under section 213 no set-off is available both because the liability under section 213 cannot arise until the company goes into liquidation and also because set-off is not available in cases of malfeasance: Manson v Smith [1997] 2 BCLC 161. In addition, the court has power to direct that the pari passu principle is to be disapplied as between the company and the person liable under section 213; Insolvency Act 215 (4).

As far as the first of these points is concerned, there is, I think, some force in it. But even on Mr Scorey’s formulation of the persons who fall within section 213 there may be some who are not fiduciaries in any real sense. So far as the second is concerned, it will apply to anyone who falls within the scope of section 213 and it does not, in my judgment, provide a basis for adopting a narrow interpretation of that section.

I should make it clear, however, that nothing I say must be taken as setting the outer limits of the scope of section 213. All that we are asked to decide is whether a person cannot fall within the scope of section 213 unless he has a controlling or managerial function within the company. Whether an “outsider” can be said to be party to the carrying on by a company of a fraudulent business may well be a question of fact and degree which requires careful analysis. That question does not arise in this case, because of the settlement agreement.

In concluding his article, Mr Foxton says:

I consider that the last sentence is right. I would dismiss this appeal.

Nathanael and Inline differ from the other claimants because there was a substantial period during which both were dissolved before being restored to the register.

In the case of Nathanael, it was struck off and dissolved on 1 February 2011, and restored to the register on the petition of HMRC on 19 March 2012; at which time a winding up order was made. Liquidators were appointed on 19 August 2013.

In the case of Inline it was struck off and dissolved on 7 December 2010 and restored to the register on the petition of HMRC on 8 June 2015; at which time a winding up order was made. Liquidators were appointed on 8 June 2015.

The significance of those facts is potentially two-fold. First, they mean that in the case of those two companies, the liquidator’s cause of action did not arise until the company in question was restored to the register and a winding up order made. That is why those claims under section 213 were not statute barred. But the companies themselves (as opposed to their liquidators) also made claims in dishonest assistance in a breach of fiduciary duty; and, in connection with those claims, relied on section 32 of the Limitation Act 1980 to postpone the running of time. The effect of restoration to the register is relevant to those claims.

Where a company has been restored to the register by court order after being dissolved section 1032 of the Companies Act 2006 provides:

The thrust of the claimants’ argument as presented in the skeleton argument is that the effect of section 1032 (1) is that the relevant companies are deemed to have continued in existence with the same directors who were in office at the date of dissolution. This consequence was not fact-dependent, but was the proper interpretation of the deeming provision. Applying that interpretation to the facts of this case, the directors in office at the date of dissolution were the wrong-doing directors, whose knowledge of the wrongdoing could not be attributed to the relevant company. It follows that during the period between the dissolution and the restoration, the company could not have discovered the fraud.

But in his oral submissions, Mr Parker submitted in the alternative that no assumption was required to be made about the existence or attributes of any director. In consequence the judge should have found that because, in reality, there was no company in existence and no directors, there was no one who could, with reasonable diligence, have discovered the fraud. This was, it seems, a point taken before the judge. It was mentioned in the grounds of appeal, but almost disavowed in the skeleton argument. Indeed, TFS prepared its skeleton argument on the basis that this argument had been abandoned. Nevertheless, Mr Scorey did not object to this argument being advanced.

The judge rejected those arguments. He held that it would be contrary to the scheme of the Limitation Act for the time that a restored company spends in enforced non-existence not to count towards the running of time for the purposes of limitation. The only difficulty lay in identifying the person or persons who would have been in control of the company during its enforced non-existence. The judge’s answer was that the deeming provision in section 1032 required the assumption that Nathanael and Inline were properly constituted companies during the period of their non-existence and that they had the minimum number of ordinarily competent directors in place. He concluded, therefore, that as in the case of the claims by other claimants, the claims by Nathanael and Inline were also statute barred.

The Supreme Court authoritatively stated the approach to statutory deeming provisions in Fowler v HMRC [2020] UKSC 22, [2020] 1 WLR 2227 at [27]:

The fifth of these principles deserves some emphasis. Although you must not permit your imagination to boggle at the inevitable corollaries of the deemed state of affairs, it does not permit a wide-ranging inquiry into consequences that were merely probable or possible consequences (see, for example, Secretary of State for Transport v Curzon Park Ltd [2021] EWCA Civ 651; [2021] PTSR 1560 at [42]).

That, in my judgment, is the short answer to the claimants’ first argument. It is not an inevitable consequence of the deeming provision that the directors in office at the date of the dissolution would have remained in office during the whole of the period of what the judge called the companies’ enforced period of non-existence if no dissolution had taken place. It is no more than a possible one. In any given case the relevant directors might have died, or become bankrupt, or might have been disqualified from acting as directors. Or, if the companies had remained in existence, HMRC might have presented winding up petitions and appointed liquidators earlier than they did. Nor is it an inevitable consequence of the companies’ deemed existence that the wrongdoing directors would have persisted in their wrongdoing. Again, it is no more than a possible one. In my judgment, therefore, the conclusion that (irrespective of the facts of any particular case) section 1032 (1) requires the assumption that the directors in office at the date of dissolution remained in office throughout the period of enforced non-existence is wrong.

Equally, I would not subscribe to the judge’s view that section 1032 (1) required him to assume that there was in place “the minimum number of ordinarily competent directors”. That, in my view, is an assumption that the statute does not require, and which is not an inevitable consequence of the required assumption.

I consider that this approach is consistent with the authorities. But in order to explain them, I need to set out shortly how the statutory framework has developed. A detailed exposition can be found in the judgments of Munby LJ in Joddrell v Peaktone Ltd [2012] EWCA Civ 1035, [2013] 1 WLR 784 and Newey LJ in Bridgehouse (Bradford No 2) Ltd v BAE Systems plc [2020] EWCA Civ 759, [2020] Bus LR 2025. In Joddrell Munby LJ explained:

Since the coming into force of the Companies Act 2006, there have been two routes to restoration; namely administrative restoration (sections 1024 to 1028); and restoration by order of the court (sections 1029 to 1032). But importantly the deeming provision on restoration applies in both cases.

In Tyman’s Ltd v Craven [1952] 2 QB 100 a company purported to apply to the county court for the grant of a new tenancy during the period of its dissolution. The hearing was adjourned, and before the resumed hearing the company was restored to the register. This court held by a majority that the effect of the restoration under section 353 of the Companies Act 1948 was that the application was to validate the application retrospectively. The majority (Sir Raymond Evershed MR and Hodson LJ) held that the effect of the restoration was to achieve an “as-you-were position” between the company and third parties. But that, as it seems to me, goes no further than to say that it is a necessary consequence of deeming the continued existence of the company that acts purportedly carried out by the dissolved company during the period of its dissolution must be treated as having been validly done. Jenkins LJ, in the minority, dissented on the ground that it was not a necessary consequence of the statutory hypothesis that the application for the new tenancy be validated, because the company could have continued in existence without making such an application.

In Joddrell v Peaktone Ltd the claimant purported to bring a claim for damages for personal injury against a company which had been dissolved. On learning of the dissolution he applied for the restoration of the company to the register. This court held that the restoration of the company retrospectively validated the action purportedly commenced against the company during the period of its dissolution. As Munby LJ put it at [29]:

In essence, the decision in that case was the obverse of that in Tyman’s case. Acts purportedly carried out (in this case against rather than by the company) were retrospectively validated.

In Bridgehouse (Bradford No 2) Ltd v BAE Systems plc [2019] EWHC 1768 (Comm), [2020] Bus LR 306 a company had been struck off and dissolved. During the period of its dissolution a counterparty purported to terminate a contract on the ground that the striking off was an event of default. The company was subsequently restored to the register. Cockerill J held that the termination of the contract was nevertheless valid. As she put it at [109], the authorities:

At [112] she added:

She concluded:

Cockerill J’s conclusion reflects my own; and, in particular that it is only the inevitable consequences of restoration that count.

A multiplicity of issues arose in Davies v Ford [2020] EWHC 686 (Ch). One of them was whether the effect of the restoration of a company was that the directors who were in office at the date of the dissolution were deemed to have continued in office during the period of dissolution. Mr Adam Johnson QC (sitting as a judge of the High Court) held that the answer was “no”. He gave a number of reasons for his conclusion:

I consider that these reasons are compelling. Indeed the crux of the claimants’ first argument under this head depends on the proposition that a company and its directors are separate legal persons with the consequence that, on the facts of this case, the knowledge of the wrongdoing directors cannot be attributed to the companies. But the conclusion that they seek to draw about the effect of the deeming provision in section 1032 (1) blurs (or even obliterates) that well established distinction.

In particular, I agree with Mr Johnson QC that any potential injustice in that conclusion is readily capable of being prevented by an application for a direction under section 1032 (3). That, to my mind, is clear from the purpose identified in section 1032 (3); namely “for placing the company and all other persons in the same position (as nearly as may be) as if the company had not been dissolved”. It is through the application of section 1032 (3) that questions of limitation are usually addressed: County Leasing Asset Management Ltd v Hawkes [2015] EWCA Civ 1251, [2016] BCC 102 at [3]; Bridgehouse (Bradford No 2) Ltd v BAE Systems plc (CA) at [26]. It is in this way that the court can balance the prejudice to the company in not having been able to begin proceedings and the prejudice to third parties who might be deprived of what would otherwise be a limitation defence.

The starting point here is that time runs against a dissolved company in the same way as it runs against an extant company: County Leasing at [30]. The court has power under section 1032 (3) to direct an extension of the limitation period in favour of a company which has a potential claim; but that power is exercisable only where the dissolution of the company was the real cause of its being disabled from pursuing its claim. If the court concludes that it would probably have failed to pursue its claim in time anyway, then the causative prerequisite implied in section 1032 (3) is missing: County Leasing at [30]. It is not simply a question of putting back the clock to the date of dissolution, but of requiring the court to ask what would have happened thereafter, during the period of dissolution, if dissolution had not occurred: County Leasing at [33].

The answer to that question is not, in my judgment, a matter of speculation or assertion: it is a question of evidence, to be decided by the court on the balance of probabilities: Davy v Pickering [2017] EWCA Civ 30, [2017] Bus LR 1239 at [60] and [71].

If the company can show that it is probable that, but for the dissolution, it would have begun proceedings in time, the court has a discretion whether to make a direction under section 1032 (3). However, considerations of fairness will generally require that the company, like any other claimant faced with a limitation defence, should be left to meet that defence by recourse to the statutory regime in the Limitation Act, rather than by a direction under section 1032 (3): Regent Leisuretime Ltd v NatWest Finance Ltd [2003] EWCA Civ 391, [2003] BCC 587 at [90].

Equally, as held in Davy v Pickering, where the limitation direction is sought by a potential claimant against a restored company, that potential claimant must show a clear causal link between the company’s dissolution and his failure to bring proceedings within the applicable limitation period. No distinction is drawn between the company and third parties. The question is whether the claimant has proved, on the balance of probabilities, that if the company had remained in existence he would have brought his claim in time. That is a question of fact, to be decided on the evidence.

In addition, it is clear that a limitation direction under section 1032 (3) should not place a person in a better position than they would have been if there had been no dissolution: Davy v Pickering at [40].

The alternative way of putting the case is that there were no directors during the period of the company’s non-existence. But this, too, seems to me to require a positive assumption to be made which the section does not require. The section requires an assumption to be made about the company, not about the absence or presence of directors. In addition, it must be firmly borne in mind that the context in which the question arises is the postponement of the limitation period under section 32 of the Limitation Act 1980. Where the claimant relies on that section, the burden lies on him to prove on the facts that he could not with reasonable diligence have discovered the fraud. That is a question of fact. The claimants failed to discharge that burden.

In my judgment, the approach of the court in relation to the making of a direction under section 1032(3) should also inform the approach to the interpretation and consequences of section 1032 (1). The three particular points are: first the company’s dissolution must have been the real cause of the company being unable to pursue its claim (County Leasing); second, the company should not be in a better position under section 1032 than it would have been if it had not been dissolved; and third what would have happened if the company had remained in existence is a question of fact (Davy v Pickering). These are all questions to be decided on the evidence, and not on legal assumptions.

It does not appear to me that at trial the claimants attempted to advance the evidential case that would have demonstrated what would have happened if the relevant companies had not been dissolved; and that would have discharged the burden of proof. That, to my mind, also disposes of the claimants’ secondary complaint under this head. They assert that the judge ought to have considered whether an ordinarily competent director of Nathanael and Inline respectively could have discovered the fraud at the same time as the liquidators could have done. The argument is that such a director did not have the investigatory powers possessed by an office holder. The difficulty with this argument is that the evidential case was not made; and the skeleton argument in relation to this ground of appeal deals with the argument in a single paragraph. The Appellant’s Notice does not ask for a new trial on this point. It would be quite impossible, in my judgment, for this court to make any findings of fact that would be necessary to sustain this argument.

I would dismiss this appeal.

The claim form was issued against TFS on 8 November 2017. The trial was due to begin in January 2021, but it was adjourned because of the illness of a crucial witness. But on 21 February 2022 the parties entered into a confidential settlement agreement which, among other things, provided for the judge to determine the substantive issues relating to the limitation defence and section 213, and also the costs of the whole proceedings. It is important to note that in the settlement agreement TFS did not admit liability; and that it agreed to pay a proportion of the amount claimed by each claimant if it were established that any of the claims succeeded either on the limitation issue or the section 213 issue. In the event, three claimants (Bilta, Weston and Vehement) failed to establish any valid claim and thus obtained nothing under the settlement agreement. Nathanael and Inline’s claim for dishonest assistance in a breach of trust failed on limitation under section 32 of the Limitation Act 1980; but succeeded under section 213.

The judge had been the “docketed judge” in relation to the action, and had dealt with various matters of case management. He also, of course, heard the trial of the issues that the settlement agreement did not resolve. The judge approached the question of costs by saying that a bright line needed to be drawn between the costs incurred before the settlement agreement and the costs incurred subsequently. The former constituted the vast bulk of the costs. It is important to note that, as the judge recorded at [9] of the costs judgment, neither side argued that the usual rule (costs follow the event) should be applied. Each party argued for an issues-based costs order.

The judge rejected the claimants’ argument that they were the successful parties because they had recovered a substantial amount of money. He considered that it was a case in which he could not form a view about who would have won if the whole case had gone to trial. The payment that TFS agreed to make (without admission of liability) arose by way of settlement, not by a finding of liability and provided no element of vindication or indication of outcome. He concluded at [20]:

He therefore decided that this was not a case in which costs should follow the event. He then considered whether he should make an issues-based order, by “reading across” the parties’ success or failure on the issues which he did decide in his substantive judgment. He considered that it was wrong to allow the issues on which there had been success or failure to inform the issues which could not be assessed one way or the other. In relation to the pre-settlement costs, he concluded at [28]:

He made an adjustment to that, by allowing the claimants 7.5% per cent of their pre-settlement costs to reflect the fact that, at TFS’s behest, there had been two adjournments of the trial.

As regards the post-settlement costs he made an order that TFS should have its costs of the limitation defence, and the successful claimants should have their costs of the section 213 issue.

There is no appeal against the judge’s decision as regards post-settlement costs. The judge himself gave permission to appeal against his decision on pre-settlement costs.

It goes almost without saying that a first instance judge has a very wide discretion (a) whether to make any order for costs at all and (b) if so, what order to make. An appeal court can only interfere on very limited grounds. As Wilson J (sitting in this court) explained in SCT Finance v Bolton [2002] EWCA Civ 56 at [1]:

The question is not, therefore, whether we would have made the same order for costs as the judge, nor even whether we disagree with it. The question is whether the judge exceeded the wide discretion that the law gives him.

Moreover, where the first instance judge has determined costs in a case in which the parties have settled their dispute (apart from costs) an appeal court is even more reluctant to interfere.

In a case in which the judge decides costs after settlement it has been said that an appeal court will only interfere if the judge’s decision is “manifestly unjust”. That test was first proposed by Mummery LJ (with whom Brooke LJ expressly agreed) in BCT Software Solutions Ltd v C Brewer & Sons Ltd [2003] EWCA Civ 939, [2004] FSR 9 at [9]. The appeal in that case was dismissed because the judge’s order was “not manifestly unjust in all the circumstances”: [20]. That test was approved by this court in Venture Finance plc v Mead [2005] EWCA Civ 325 at [25] and [34]; Promar International Ltd v Clarke [2006] EWCA Civ 332 at [35] and Bray v Bishop [2009] EWCA Civ 768 at [5]. In Patience v Tanner [2016] EWCA Civ 158 at [33] Gross LJ questioned whether the reference to “manifest injustice” imposed a further threshold requirement, but was content to proceed as if it did. But as David Richards LJ pointed out in Powles v Reeves [2016] EWCA Civ 1375, [2017] 1 Costs LR 19 at [25] the decisions in Venture Finance plc v Mead and Bray v Bishop do not appear to have been cited in Patience v Tanner. In Powles v Reeves David Richards LJ confirmed at [26]:

The claimants do not criticise the judge for having distinguished so sharply between pre-settlement and post-settlement costs. But they argue that the judge did indeed make an error of principle in relation to the pre-settlement costs. They say that because the outcome of the litigation was that TFS ended up “writing the cheque” the claimants were the successful party. Even though the judge might not have been able to decide who would have won on liability if the whole case had gone to trial, nevertheless this was a case where the terms of the settlement agreement contained the answer.

In the BCT case, for example, Mummery LJ said at [7]:

In R (M) v Croydon LBC [2012] EWCA Civ 595, [2012] 1 WLR 2607 Lord Neuberger MR differentiated between cases where the claimant had achieved all the relief that he sought, and cases in which the claimant had achieved part of the relief he sought. In the latter kind of case, he said at [50]:

He added at [62]:

Gossage v Bishton [2012] EWCA Civ 717 is another example of a case in which the achievement of partial success in a compromise resulted in no order for costs.

Mr Parker also relied strongly on the decision of this court in McKeown v Langer [2021] EWCA Civ 1792, [2022] 1 WLR 1255. That was a case of a petition for unfair prejudice brought under section 994 of the Companies Act 2006. There was a split trial of liability which was to be followed by the question of valuation if a buy-out order was made. The judge hearing the trial made an order for costs. The argument raised on appeal was that since the respondent had made a global Calderbank offer the judge was wrong not to have deferred the question of costs until after the second trial. This court rejected that argument. Whether to make a costs order was within the discretion of the judge. It is perfectly true that at [37] Green LJ said that:

But that was in the context of a split trial; and the actual decision in that case was that the costs order that the judge in fact made was within the ambit of his discretion. The court did not say that he would have been wrong to have approached the question of costs differently. In my judgment, Mr Parker’s reliance on that case seeks to confine a judge’s discretion over costs impermissibly.

Here, the claimants say, the action was for the recovery of money, and the claimants recovered a substantial sum. The judge should have been able to tell from the size of the settlement sum how the costs should be borne. It was not necessary (and was an error of principle) for him to have considered who might have won at trial if all the issues had been litigated. The claimants therefore argue that they should recover all their costs down to the date of the settlement agreement. Although I consider that there is considerable force in the argument that the terms of the settlement may in some cases provide a good proxy for deciding who is the successful party without having to second guess the result of a contested trial, this is not, in my judgment, a case of complete success, but of partial success. In addition it is, I think, necessary to consider the judge’s costs order in the round in order to decide whether it was “manifestly unjust”.

The result of the claimants’ argument would produce the result that three claimants (Bilta, Weston and Vehement) whose claims failed completely would recover the bulk of their costs. That proposition only has to be stated in order to be seen to be unjust. This was not a case in which all costs were common costs. Substantial costs were incurred by TFS on individual claims by the individual companies. In addition, in relation to all the claimants, TFS won on a very substantial issue, namely the limitation defence. That entailed a detailed examination of the facts (unlike the section 213 issue which was a question of law). It could not be right that the claimants should recover their pre-settlement costs of that issue on which they all failed. Mr Parker was willing to accept that the claimants should not recover the costs of preparing Mr Hellard’s witness statement on that issue, but there would no doubt have been other costs incurred in relation to that issue (for example, statements of case, disclosure and inspection of documents). There would also have been costs incurred on the question of quantum which were overtaken by the settlement agreement. If one were to apply the principle that the unsuccessful party pays the successful party’s costs, and did not disaggregate the individual claimants, it might well be that TFS would recover three fifths of their costs, and Nathanael and Inline (or their liquidators) would recover all their costs, apart from the costs (both pre settlement and post-settlement) of the limitation issue. Although the judge did not approach it in quite this way, netting those costs off, one can see how the judge could have arrived at the decision he did.

As Lord Neuberger said in Croydon, in cases of partial success it may not be easy to assess whether the defendants should have their costs of the issue on which the claimant did not succeed, and whether that would wipe out the costs which the claimant might recover in relation to the issue on which he won.

It is also worth examining the BCT case a little more closely. BCT was a claim for copyright infringement. Shortly after the trial began the parties settled on terms that BCT should recover £10,000. The judge was invited to determine costs. He made a complex costs order to reflect the fact that BCT had radically amended (and narrowed) its claim a month before trial. The only part of his order that was appealed was his decision that BCT should pay 62.5 per cent of Brewer’s costs after the amendment. This was, then, a case in which BCT did recover a sum from Brewer. At [18] Mummery LJ said that if he had been the judge, he could probably have made no order for costs. Chadwick LJ said at [27] that he agreed with Mummery LJ that the judge could not have been criticised if he had decided to make no order. Accordingly, even where the claimant does recover money from the defendant, a judge is not necessarily wrong in making no order for costs.

The Venture Finance case was a claim under a guarantee against two guarantors. The amount claimed was just under £300,000. Shortly before the claim was due to be heard the two guarantors agreed to pay £92,500 each. The judge made an order for costs. Although this court made an order for costs (which differed from the judge’s order), Chadwick LJ said at [12]:

Capita (Banstead 2011) Ltd v RFIB Group Ltd [2017] EWCA Civ 1032, [2017] 4 Costs LR 669 is another case in which despite the fact that the claimants recovered a substantial sum the judge was entitled to make no order for costs, in view of the fact that they lost on one important issue. Thus even in a case where the claim was for money, and the claimant had recovered a substantial sum, it would not have been an error of principle for the judge to have declined to make an order for costs. The mere fact that one party pays money to another does not necessarily mean that the receiving party is the successful party, particularly in a case of a group of claimants pursuing individual claims, some of whom succeed and some of whom fail: Atlasjet Havacilik Anonim Sirketi v Kupeli [2018] EWCA Civ 1264, [2019] 1 WLR 1235.

In my judgment, the judge’s decision on pre-settlement costs did not fall outside the generous ambit of his discretion. I would dismiss this appeal.

I would dismiss all the appeals.

I agree.

I also agree.