Holland v Revenue and Customs & Anor

This appeal by Mr Michael Holland is against an order dated 4 July 2008 made by Mr Mark Cawson QC sitting as a Deputy High Court Judge of the Chancery Division, his order following his 80-page judgment of 24 June 2008. The judgment followed a ten day trial of 42 originating applications issued by The Commissioners for Her Majesty’s Revenue and Customs (‘HMRC’) on 27 July 2006 against Mr Holland and his wife Linda. The applications, made under section 212 of the Insolvency Act 1986, were brought on the basis of allegations that, as de facto directors of 42 insolvent companies, Mr and Mrs Holland had been guilty of misfeasance and breaches of duty in causing the payment between 24 April 2002 and 19 October 2004 of unlawful dividends totalling some £13m.

I record at the outset my tribute to the care and thoroughness of the judge’s judgment (reported at [2008] 2 BCLC 613; [2009] Bus L.R. 1). Its outcome was that he dismissed the claims against Mrs Holland. As for Mr Holland, the judge focused on three different periods. First, in respect of the dividends paid during the period 24 April 2002 to 18 August 2004, the judge held that Mr Holland was at no stage liable, or if he was, that he ought to be relieved from liability pursuant to section 727 of the Companies Act 1985. Second, in respect of the dividends paid during the few days from 19 to 22 August 2004, the judge relieved Mr Holland from liability under section 727. Third, in respect of the dividends paid during the period 23 August to 19 October 2004, the judge declared that Mr Holland had been guilty of misfeasance and breach of duty in causing their payment. He ordered an assessment of the contributions to the companies’ assets that Mr Holland must pay the liquidators, limiting them to ‘the amount of Higher Rate Corporation Tax that the Composite Companies failed to provide for that accrued due in respect of trading in the period 23 August 2004 to 19 October 2004 together with interest thereon’ at a rate to be determined. The amount of such tax was about £144,000. He ordered Mr Holland to pay half of HMRC’s costs, not including those of their claim against Mrs Holland.

With the judge’s permission, Mr Holland appeals on four grounds; and HMRC cross appeal on two. Mr Peter Knox QC appeared before us, as below, for Mr Holland. He led Mr Aidan Casey, who represented Mrs Holland at the trial. Mr Michael Green appeared before us, as below, for HMRC. Mrs Holland, formally a respondent to her husband’s appeal, was not represented.

The litigation arose out of the setting up by Mr and Mrs Holland in 1999 of a complicated structure of companies whose business was the administering of the business and tax affairs of contractors working in various sectors, mainly IT. Each contractor was taken on as an employee of one of the companies and allotted a non-voting share. This enabled him to be rewarded on a weekly or monthly basis by way of both salary and dividends. The contractor’s services were provided to clients through an agency which paid the relevant company. It was the essence of the basis on which the companies were established that each would be liable to pay corporation tax at only the small companies’ rate, which was between 19% and 21% at the material time. As long as the companies were not regarded as ‘associated’ for the purposes of the relevant tax legislation, they could achieve this provided that each kept its annual profits below the £300,000 threshold, as it did. Otherwise each company would be liable to pay what the judge called higher rate corporation tax (‘HRCT’), that is tax at the main corporation tax rate, which was between 30% and 33% at the material time. I will, like the judge, refer to tax at that rate as ‘HRCT’.

The scheme was flawed, with the consequence that each company was liable to pay HRCT throughout its trading life. They had, however, made no provision for it; and each had, throughout such life, declared and paid dividends which should not have been paid because there were insufficient distributable reserves to permit them. All the companies stopped trading on 13 October 2004, they went into administration on 19 October 2004 and into creditors’ voluntary liquidation in February 2005.

Mr Holland was not a de jure director of any of the companies: the sole director of each was a company of which he was a director. But the judge found that he was a de facto director of each company and so was answerable to HMRC’s claim under section 212. Mr Holland challenges that finding as wrong. If he is right on that, that is the end of HMRC’s case. If he is wrong, he argues further that he was anyway not, as the judge held, liable to reimburse the companies in respect of the dividends unlawfully paid during the period 23 August to 19 October 2004; alternatively, that the judge misdirected himself in refusing him relief under section 727 of the Companies Act 1985 against the claim in respect of the dividends so paid; and that the judge was also in error in the manner in which he determined the contribution to the companies’ assets that Mr Holland should make, the assertion being that his order requiring a contribution of some £144,000 over-compensated the companies.

HMRC, by their cross-appeal, make common cause with Mr Holland that the judge misdirected himself as to the contribution that Mr Holland should make to the companies’ assets, but say that his error led to an order that under-compensated the companies. They say he should have ordered Mr Holland to repay the full amount of the unlawful dividends paid during the period 23 August to 19 October 2004 (about £1.3m). They also say that the judge was wrong to let him off liability in respect of the four-day period from 19 to 22 August 2004 inclusive during which unlawful dividends of some £140,000 were paid.

From about June 1997 to 1999 Mr and Mrs Holland ran a company called Paycheck Services Limited (‘Paycheck’). Paycheck’s function, in return for a fee, was to administer the business and taxation affairs of contractors working in various sectors. Each contractor became an employee of Paycheck and was allotted a non-voting share entitling him to dividends as well as a salary. Most contractors did not pay higher rate income tax and the bulk of their income from Paycheck was by way of a dividend. By late 1998 it appeared that Paycheck’s profits were likely to exceed the £300,000 threshold above which they were liable to be charged HRCT. Mr and Mrs Holland wanted to expand their business whilst avoiding liability to corporation tax at that rate. To that end, on the advice of solicitors (Mr Carl Newton of Neil Myerson), counsel (Mr Nigel Ginniff) and chartered accountants (Mr Matravers of Matravers & Co), they established a new company structure in 1999. It operated until 13 October 2004, with all the companies then going into administration on 19 October 2004. The structure was as follows. My description is gratefully derived directly from the judge’s.

Mr and Mrs Holland each held 50% of the issued shares in, and were directors of, a trading company they operated called Paycheck Services Limited (not the original Paycheck, a new company) (‘Paycheck Services’). Paycheck Services held 100% of the issued shares and Mr and Mrs Holland were directors of Paycheck (Directors Services) Limited (‘Paycheck Directors’) and Paycheck (Secretarial Services) Limited (‘Paycheck Secretarial’).

Paycheck Directors and Paycheck Secretarial were incorporated to act respectively as the sole director and secretary of each of the 42 trading companies which the judge called, as will I, ‘the composite companies’. They all had similar names distinguished only by a number: Paycheck Services 3 Limited, Paycheck Services 4 Limited and so on. Each company had one voting ‘A’ share and some 50 non-voting shares, each non-voting share being in a separate class (B1, B2, C1, C2 and so on). The ‘A’ share of each company was held by Paycheck Services Trustee Limited (‘PST’), a company in which Mr and Mrs Holland each held 50% of the issued shares and of which they were directors. The non-voting shares of each company were held by some 50 employees of the company, each employee holding one each of the separate classes of shares. Under the terms of a trust deed, PST held the ‘A’ share of each company for the benefit of the shareholder/employees (i.e. holders of the non-voting shares) of that company.

Article 8(b) of the composite companies’ Articles of Association provided:

Article 12(b) specified that the minimum number of directors should be one, and that whilst there was only one director the sole director should have authority to exercise all the powers and discretion by Table A and the Articles expressed to be vested in the directors generally.

The purpose behind the establishment of the new structure was to provide to the shareholder/employees of each composite company the same tax advantages they would have enjoyed had each set up his own company: in particular, the benefit of a corporation tax liability on profits at only the small companies’ rate. In addition, they enjoyed the benefit of Paycheck Services’ administrative support and were thereby relieved of the burden of having to operate their own company. The composite companies contracted out the services of each of their shareholder/employees (typically through employment agencies); and out of the income received in respect of each employee/shareholder, they paid: (i) a fee to Paycheck Services for its administrative services; (ii) a salary to each shareholder/employee, typically limited to the national minimum wage and the associated PAYE tax and National Insurance contributions; and (iii) after making provision for the payment of corporation tax at the small companies rate – not HRCT – a dividend to each shareholder/employee.

Such dividends were paid on a regular basis. The shareholder/employees would put in time sheets for their work done, the figures were entered into Paycheck Services’ computer and the accountancy software on it calculated the dividend payable. The program would generate a document purporting to be a minute of a board meeting of the relevant composite company, recording as present ‘M. Holland Paycheck (Director Services) Limited, L.M. Holland Paycheck (Secretarial Services) Limited’ and resolving that a dividend of a specific amount be distributed to the specified shareholder/employee. The computer generated on the minute a copy of Mr Holland’s signature, beneath which appeared ‘for and on behalf of Paycheck (Director Services) Limited.’

Each composite company, through the computer program operated by Paycheck Services, maintained in relation to each shareholder/employee a ‘mini profit and loss account’. That recorded the figures behind the dividend payments and, taken collectively, the accounts in respect of each shareholder/employee of a particular composite company allowed the ascertainment of the overall financial position of the company. Each company’s annual accounts were prepared by Mr Rees of Speechleys, chartered accountants. Mr Tony Ashton kept the records and summaries of the mini profit and loss account for each shareholder/employee from which he prepared a balance sheet and profit and loss account.

At the relevant time, corporation tax at the small companies’ rate was payable up to a profits threshold of £300,000 a year. There was only a tax advantage in being a shareholder/employee of a composite company if the company was not liable to pay HRCT: tax liability at that rate would reduce the level of dividends payable as compared with the level payable if the tax liability was at the lower rate. The structure was set up on the intended basis that corporation tax would only be payable at the lower rate. There was, however, in this respect a flaw in the structure that was not spotted when it was established.

It is unnecessary to explain in detail how the flaw arose. All I need say is that the combined effect of (i) sections 13 and 416 of the Income and Corporation Taxes Act 1988 and (ii) PST’s control of each of the composite companies by its holding of the ‘A’ shares, meant (iii) that the composite companies were ‘associated’, with the consequence that their collective profits had to be aggregated; and as they collectively exceeded the £300,000 threshold, the result was each would be liable to HRCT even though the profits of each viewed separately were below that threshold. All this was in fact foreseen when the structure was established. It was, however, believed that an escape from consequence (iii) was provided by Extra Statutory Concession C9 (‘ESC C9’), which fathered the thought that the companies would not be regarded as ‘associated’. It eventually emerged, however, although not until June 2004, that section 417(3) of ICTA defeated the claimed reliance on ESC C9: the problem was that, as Mr Holland was the settlor of each of the trusts of the ‘A’ shares, he was regarded as in control of all the composite companies, which took the case outside ESC C9. The consequence was that, because in the meantime each composite company had declared dividends after providing only for corporation tax at the small companies’ rate, each faced a liability for HRCT on its profits for prior periods for which no provision had been made. The result was that, during the companies’ trading lives, dividends of some £13m were paid that should not have been paid and that when the companies went into administration on 19 October 2004 they had a deficiency as regards creditors of some £3.5m entirely represented by unpaid corporation tax.

When the new structure was set up in 1999 Mr and Mrs Holland believed, or hoped, on the basis of professional advice, that ESC C9 would save each composite company from a liability to HRCT provided it kept its annual profits below £300,000, as each did: that is because it would not be regarded as ‘associated’ with the other composite companies. They did not, however, receive unqualified advice that this result would be achieved.

Mr Newton (of Neil Myerson) was clear from the outset that it was essential that the composite companies should not be regarded as ‘associated’. He believed that ESC C9 would provide the answer but overlooked the fatal effect of section 417(3) of ICTA by reason of Mr Holland’s position as common settlor. He sought the advice of Mr Ginniff (counsel) on the structure and on the ‘association’ point. Mr Ginniff’s advice (faxed to Mr Holland on 22 January 1999) contained some contradictions, but gave encouragement, if less than total, that the companies would not be regarded as associated.

Mr Newton wrote a letter of advice to Mr and Mrs Holland on 2 February 1999. He referred to the advice of counsel, who had since also discussed the structure with an Inspector of Taxes. Mr Newton’s advice was that none of the professionals could guarantee that HMRC would not try and attack the structure ‘with a view to trying to tax the whole of the profits made by each Composite Company and Paycheck Services as one.’ He advised restricting the profits of each company and of Paycheck Services to £150,000.

The new structure was established in February 1999. It was one of about 100 similar such structures in the United Kingdom. The composite companies were established over a period extending to about July 2000.

Between November 2000 and February 2001 Speechleys submitted to HMRC accounts and corporation tax returns for the composite companies. On 21 March 2001 Mr Watson of HMRC wrote to them in relation to companies 7, 8 and 10 asking for a detailed profit and loss account and saying he expected the notes to indicate if the companies were grouped or associated. Mr Rees replied on 26 March in relation to the three companies, saying that as they were not grouped or associated he failed to see the need for a note to this effect on the accounts. He followed that on 30 March with a letter explaining the company structure, including that ‘the “A” share in each company was held by a trust set up by Mr and Mrs Holland’, so drawing HMRC’s attention to the ‘common settlor’ point.

On 27 April Mr Watson wrote to Mr Rees, referring to his letter of 26 March but not expressly to that of 30 March. Mr Rees’s evidence was that he read Mr Watson’s letter as a response to both letters and as accepting that the three companies were not relevantly associated. The judge found that Mr Watson’s letter was intended to be written in response to both and that it:

Following Mr Watson’s letter of 27 April, there was silence on the part of HMRC until in February 2002 Mr Williams wrote to Paycheck Services and the composite companies saying that HMRC intended to enquire into their tax returns, suggesting that the companies were or might be associated and asking why HRCT had not been paid. Speechleys’ response was that HMRC had agreed in April 2001 that they were not associated. Mr Williams’ rejoinder on 24 April 2002 was that he proposed to research the matter. His letter was important, since it was only from its date that HMRC claimed in the proceedings that Mr and Mrs Holland ought to be found liable for permitting the dividends to be paid without making provision for HRCT. Mr Rees told Mr Holland of Mr Williams’ stance and advised him that the likely outcome was that Mr Williams would adopt the same approach as Mr Watson had.

During the rest of 2002 there was sporadic and inconclusive correspondence between Mr Rees and HMRC on the topic, including as to the application to the structure of ESC C9. Mr Williams did not put his finger on the defect in the structure. There was little advance during the first 11 months of 2003.

December 2003, however, marked the beginning of the end. On 4 December Mr McGillivray of HMRC wrote that they were opening formal enquiries into the claim for marginal tax relief made by all the composite companies in their tax returns for the year ended 31 July 2002. He advised Mr Rees that a single ‘covering’ letter of appeal against HMRC’s stance would suffice and on 22 December Mr Rees wrote a form of letter of appeal asking for ‘an appeal meeting before the Special Commissioners to make a definitive ruling in this matter’ (i.e. whether or not the companies were associated).

By 7 January 2004 Mr Rees had learnt - and told Mr Newton who told Mr Holland - that HMRC had decided to attack the structure on the basis that the composite companies were associated for tax purposes. Mr Rees advised Mr Newton that the HRCT, if payable, exceeded £2m and the judge found that Mr Holland knew the amounts involved. He also found that Mr Rees continued to believe that the companies had a good case and that he would have conveyed that to Mr Holland. At a meeting on 24 February Mr Rees advised Mr and Mrs Holland that outstanding assessments claimed about £2m additional tax. It was agreed that if the appeals were lost, the companies must enter voluntary liquidation. Mr Newton advised Mr Holland that ‘theoretically’ the liquidator could make a claim against the directors for the payment of unlawful dividends but that his honest trading of the companies in the belief that they were not associated would provide a defence. A decision was taken (which the judge found to be a ‘collective’ one) to continue trading in the normal way until 17 March, when a statement was expected as to a possible change in relation to the National Insurance contribution regime. In the event no change was proposed or made and the companies continued to trade.

The judge referred to some exchanges between Mr Rees and Mr Russell of HMRC on 28 May and found that Mr Rees genuinely believed from them that HMRC had concluded that none of the composite companies was associated and that the sole purpose of a meeting that had been fixed for 21 June was to ‘tie up loose ends’. That meeting was attended by Mr Newton, Mr Rees, Mr Holland, Mr Ashton, Mr Russell and Mr McGillivray. Whereas the evidence of Mr Newton, Mr Rees and Mr Holland was that they went into it believing that the issues had been effectively resolved, HMRC’s stance was that they remained wide open. There was discussion as to the solvency of the companies, which would be unable to pay any extra corporation tax due – amounting in total to about £1.4m per accounting period. Mr Newton’s note accurately summarised part of the discussions thus:

The crucial development came on 25 June when Mr Russell wrote a letter to Neil Myerson. For the first time, among other reasons, HMRC relied, correctly, on section 417(3) of ICTA as to why the companies were to be regarded as ‘associated’. Neil Myerson sought Mr Ginniff’s advice on 6 July, but neither appears to have picked up that key point.

Mr Ginniff settled some draft proceedings for judicial review against HMRC. Mr Newton was unhappy with the draft and, on 3 August, instructed new counsel, Mr Conrad McDonnell. On 6 August Mr McDonnell advised Mr Newton on the telephone that the section 417(3) point ‘blows our claim out of the water’, which he followed with written advice that Mr Newton received on 9 August. Mr McDonnell there gave firm advice that the proposal to bring judicial review proceedings should be abandoned. He advised that an attempt be made to bring HMRC’s enquiry to a close:

Mr McDonnell’s advice came as a shock to Mr Newton, Mr Rees and Mr Holland. There were lengthy telephone conversations on 10 August. Although the view taken was that the game was probably up and that what was required was a new, tax-effective corporate structure, a decision was made to take Leading Counsel’s advice. It was decided that, subject to the views of Leading Counsel, the judicial review claim should not be pursued. Leading Counsel was also to be asked to advise if the suggested new structure would work. There was discussion as to the possibility of a deal with HMRC, Mr Newton’s claimed recollection of the meeting of 21 June being that HMRC would forego the additional tax if the composite companies were closed down. Mr Newton thought they should also take Leading Counsel’s advice on that point. His view was that HMRC might not be prepared to forego that tax merely on the basis of a proposed restructuring; HMRC apparently wanted a dismantling of composite company structures. He advised Mr Holland that he was ‘probably now wrongfully trading’ the companies.

An urgent consultation with Mr John Tallon QC was arranged for 18 August, which Mr Holland attended. Mr Tallon essentially agreed with Mr McDonnell. He considered that HMRC had handled the company ‘association’ issues badly but that, although leave for a judicial review application might be granted, the composite companies would ultimately lose. He advised that Mr McDonnell’s suggested alternative structure was sound and would avoid ‘association’ issues. He advised an immediate response to HMRC’s letter of 25 June and settled a draft letter. The letter would propose a meeting with HMRC, at which a deal over the additional tax might be discussed. HMRC were to be told that the structure was being changed. Mr Newton pointed out at the consultation that if the existing companies continued to trade in the meantime, there was a risk of a wrongful trading charge in respect of the period following Mr McDonnell’s advice. Mr Tallon said he was not an insolvency specialist and could not advise on that.

Mr Newton gave evidence of what happened on the train journey back from that consultation. He advised Mr Holland that he and Mrs Holland might be trading wrongfully. He explained that Mr Holland should not continue to run the companies unless there was a reasonable prospect of avoiding insolvent liquidation. He advised that there might be an argument that dividends paid at that time would be unlawful ‘but that needed to be balanced against the need to also mitigate in the interest of the creditors, to mitigate the amount of creditors, and also whether or not a deal could be done with the Revenue.’ He advised Mr Holland he would have to be able to justify continuing the current operation; and that if the companies were closed at that point, there would be creditors created in terms of the clients to whom services were being provided and that ‘there would undoubtedly be breaches of contract, because men would not go to work for people if they were not paid. So this was an issue.’ He said:

The judge accepted Mr Newton’s evidence about what he said during the train journey and found also that Mr Holland was in no mood to engage in the discussion that Mr Newton initiated, which he said probably explained why Mr Holland’s evidence was that he had no recollection of it. Mr Newton did not follow up his advice with a letter to the like effect. There was no evidence, following the train journey, of either Mr Newton or Mr Rees having given further advice to Mr Holland as to the propriety of paying dividends or of Mr Holland having sought such advice on the point.

On 19 August the letter settled by Mr Tallon was sent to HMRC. On 23 August Speechleys wrote to Neil Myerson, referring to the proposal to have the new companies in place for the purposes of the new structure by 1 October. On 25 August Mr Rees wrote to Barclays, saying the existing companies would trade until 30 September and the new ones from 1 October, referring (misleadingly, the judge found) to HMRC having ‘basically given us an amnesty period during which to restructure the company along the lines discussed above,’ a statement of fiction. On the same day Mr Rees wrote to Neil Myerson, saying:

The new companies were incorporated on 3 September. On 10 September Mr Russell wrote to Neil Myerson saying the enquiry process was complete, that there were no statutory grounds for an appeal, that judicial review was the only avenue and asking how the composite companies intended to proceed. On 20 September the shareholder/employees were written to in misleading terms, giving the impression that there was simply to be a change of name rather than a transfer of their businesses to new companies under a new structure. On the same day Mr Newton spoke to Mr McGillivray with regard to arranging a settlement meeting, which was fixed for 4 October. Further advice on the new structure was received from Mr Tallon.

The 4 October meeting took place, attended by Messrs Newton, Rees, McGillivray and Russell. HMRC were informed, for the first time, of the intention to transfer the business to the new structure. The proposal was put to HMRC that they should accept that ESC C9 did apply to the existing companies down to the end of October, on the basis that they would cease to trade then, pay all outstanding corporation tax at the small companies’ rate and then be dissolved. Mr Newton made the point that if they were forced to cease trading earlier and enter an insolvency process, the pot available to HMRC would be less: it was suggested that the employee/shareholders would bring claims against the companies and there would also be the costs of the insolvency process.

HMRC rejected that proposal. By his letter of 5 October Mr Russell informed Neil Myerson that HMRC still sought HRCT from 2002. The letter showed that HMRC considered that unlawful dividends had been paid. The letter was in fact only received by Neil Myerson on 13 October, when it was forwarded to Mr Holland. Mr Newton advised Mr Holland that the directors should act in the best interests of creditors and that there was no possibility of a deal with HMRC. He advised that no further dividends should be declared as there were no distributable reserves. No further dividends were declared after 13 October.

On 19 October Mr Conn and Mr Stanley were appointed joint administrators of each of the composite companies and thereafter the various contracts were transferred to the new companies, the new companies assumed the businesses of the old and that avoided claims for breach of contract against the old companies. The seamless transition left the employee/shareholders, their clients/customers and Mr and Mrs Holland apparently content. The only loser was the composite companies’ sole creditor – HMRC, for unpaid corporation tax.

In the meantime, over the period from 18 August (the date of Mr Tallon’s advice) to 13 October, the composite companies had of course continued to trade as before. They had a total of some 500 employees, most of whom were also shareholders. They continued to be rewarded by wages and (in the case of those who were also shareholders) by weekly or monthly interim dividends: the companies continued to declare and pay unlawful dividends and continued to fail to provide for HRCT. The reason that no such provision was held back from the dividend payments was that it was perceived that, if they were, the employees would promptly leave the companies and join other groups of composite companies whose structure did not suffer from the same tax defect. The judge made no express finding to this effect, but we were told it was common ground. The companies were also subject to the burden of contracts under which their employees’ services had been contracted out to third party customers, to whom they provided a wide range of services. If, having learnt of the HRCT position on 18 August 2004, the companies had promptly stopped trading, it is said they would have been faced with claims not just by the employees but also by customers whom they would have let down.

The net effect of the continued trading from 19 August to 13 October 2004 was that during this period the composite companies, all of which were insolvent, paid some £1.3m of unlawful dividends and that the unpaid HRCT increased by about £150,000. Against this, as the judge found, they were saved from claims that would otherwise have arisen if they had stopped trading in advance of the transfer to the new companies. During this period Mr Holland acted under the guidance of his advisers, who did not advise him to stop trading. The companies entered administration with a total deficiency of some £3.5m in respect of unpaid corporation tax.

The joint administrators made no claims against Mr and Mrs Holland, nor did the liquidators who were subsequently appointed. But HMRC took a different attitude. Whilst they accepted that the only de jure director of the composite companies was Paycheck Directors (of which Mr and Mrs Holland were directors), they asserted that Mr and Mrs Holland were both de facto directors of each of the composite companies. They alleged that in breach of duty as such they had, since 25 April 2002 caused them to pay unlawful dividends contrary to sections 263 and 270 of the Companies Act 1985 because they had insufficient distributable profits with which to pay them because of the accruing liability to HRCT. The relevance of 25 April 2002 is that HMRC accepted that until then Mr and Mrs Holland might have been led by their prior correspondence to believe that the companies were not liable to HRCT. HMRC claimed that Mr and Mrs Holland were strictly liable to reimburse to the companies the full amount of the dividends wrongly paid (some £13m). But they tempered the claim to the amount of HRCT for which the companies had failed to provide since 25 April 2002, some £3.5m.

Mr and Mrs Holland achieved considerable, but not total, success in meeting the claim. The claim against Mrs Holland was dismissed. Whilst the judge held that Mr Holland was a de facto director of the composite companies, he held that he was at no stage liable in respect of the dividends paid during the period down to 18 August 2004, or if he was, he granted him relief pursuant to section 727 of the Companies Act 1985. But he held that after then, by which time he had Leading Counsel’s advice that HRCT was payable, Mr Holland knew that there were insufficient profits to justify the further distributions. The judge held that, in principle, Mr Holland was strictly liable to repay the dividends wrongly paid (about £1.3m) but that he had a discretion enabling him to temper this in two ways: (i) to allow a period of grace until and including 22 August; and (ii) by limiting the contribution he must make to the companies’ assets to the increased deficit in HRCT accruing in the period from 23 August onwards (about £144,000). Save that I understand that the period of grace was granted by way of relief under section 727, the judge otherwise declined to exonerate Mr Holland from liability under that section. But the judge considered that he could limit the ordered contribution in the way he did in exercise of the discretion conferred by section 212 of the Insolvency Act 1986.

I first set out the material parts of section 212 of the Insolvency Act 1986 under which HMRC’s applications were made:

Mr Holland was not a de jure director of any of the composite companies. There is, however, no doubt that ‘officer’ in section 212(1)(a) includes a de facto director: the definition of an ‘officer’ in relation to a body corporate in section 744 of the Companies Act 1985 includes ‘a director, manager or secretary’; and section 251 of the Insolvency Act 1986 incorporates that definition and also defines a ‘director’ as including ‘any person occupying the position of director by whatever name called’.

Mr Holland’s first ground of appeal is the so-called ‘corporate director point’. Mr Knox’s submission was that the only director of the composite companies was Paycheck Directors. Whilst Mr Holland was admittedly, with Mrs Holland, a director of Paycheck Directors and they were the human agents behind all that Paycheck Directors did in its capacity as a director of each of the composite companies, those facts did not entitle the judge to find that he was also a de facto director of those companies. That was a step too far which ignored the separate legal identities of Paycheck Directors and Mr Holland.

The judge dealt with this issue in paragraphs [164] to [179] of his judgment. He first summarised the principles applicable for determining whether an individual is or is not a de facto director of a company. I need not repeat them because Mr Holland conceded that, subject to the point that all he did in relation to the affairs of the composite companies was exclusively in his capacity as a director of their sole director, his actions would have justified his being characterised as a de facto director. But his case was that as all he did was in such capacity, there was no scope for treating him as such a director. There was, in particular, no allegation that Paycheck Directors was a sham or façade or that Mr Holland acted outside his authority as a director of Paycheck Directors (paragraph [172] of the judgment).

The judge’s view, however, was that Mr Holland’s argument did not meet the relevant point. He said the question raised by section 251 of the Insolvency Act 1986 was the essentially factual one of whether or not Mr Holland was someone ‘occupying the position of director [of the composite companies] by whatever name called’. In answer to that question, the judge ruled as follows:

Mr Knox challenged that reasoning and conclusion. Starting at the beginning, he pointed out that the Companies Act 1862 made no express provision for the appointment of corporate directors but that the issue as to whether such a director could be appointed arose in In re Bulawayo [1907] 2 Ch 458. The argument against such an appointment was, inter alia, that if it were permitted, how could all the sanctions against an errant director be properly applied? Warrington J held that nothing in the legislation prevented the appointment of a corporate director.

Mr Knox said that subsequent Companies Acts did recognise that a company can have a corporate director. The Act in force at the time material to these proceedings was the Companies Act 1985 and I agree that it recognises both that a company can have a corporate director (see, for example, section 305(4)(b)) and also that such a corporate director can be a company’s sole director, as is implicitly recognised by section 283(4)(b):

Mr Knox submitted that it follows that it is not open to the court to pay lip service to the appointment of a company as the sole director of another company and then ignore the legal effect of its appointment as such by treating it as in reality no more than a nominee, alias, agent or façade for the human beings who must inevitably act as its decision-making organ. Those human agents will have to procure the corporate director to do what it has to do in its capacity as such: there is no other way in which the corporate director can act. But the acts so procured will be the acts of the corporate director, not of its human agents. To hold otherwise is to ignore the separate legal identity of the corporate director on the one hand and its own human directors on the other (Salomon v. Salomon and Company, Limited [1897] AC 22, at 33, 34 and 42, 43). Whilst in appropriate circumstances, policy considerations may dictate that it will be just to pierce the veil of incorporation and identify the company with the individuals behind it, no-one has suggested that the facts of this case required any such piercing.

Mr Knox submitted that the judge therefore went down the wrong track in the way he dealt with this issue. His approach was that as Mr Holland was, as a matter of fact, directing the affairs of the composite companies, that was enough to render him answerable under section 212 because it meant that he was a ‘person occupying the position of a director, by whatever name called’. It mattered not that he was doing so exclusively in his capacity as a director of Paycheck Directors. Mr Knox said that accepting, as the judge did, that Mr Holland did all that he relevantly did as in his capacity as a director of Paycheck Directors, it was Paycheck Directors that was conducting the relevant direction of the composite companies; and it alone was responsible for any misfeasance or breach of duty towards the composite companies. The fixing by the judge of liability on Mr Holland by an appeal to the definition in section 251 overlooked that Mr Holland had not taken any relevant part in the management of the composite companies: their management had been conducted exclusively by Paycheck Directors.

Mr Knox submitted that the judge’s reasoning was inconsistent with that of Millett J in In re Hydrodam (Corby) Limited [1994] 2 BCLC 180. The facts of that case were that Eagle Trust plc (‘ET’) wholly owned Midland City Partnerships Ltd (‘MCP’), which wholly owned Landsaver MCP Limited, which wholly owned Hydrodam (Corby) Limited (‘HCL’). The only de jure directors of HCL were two Channel Island companies. HCL went into compulsory liquidation and its liquidator brought claims under section 214 of the Insolvency Act 1986 for wrongful trading against 14 defendants, including ET, one of its subsidiaries and all its directors. Two of those directors were Mr Thomas and Dr Hardwick, who applied for the proceedings against them to be struck out.

Millett J said, at 182, that directors may be of three kinds:

Millett J continued by explaining that it was conceded that liability under section 214 extended to de facto as well as to de jure and shadow directors, but he said that the statutory liability was ‘imposed exclusively upon directors of one or other of the three kinds that I have mentioned.’ That meant that the liquidator had to plead and prove against each defendant that he was such a director of HCL. He explained the difference between a shadow and a de facto director, saying (at 183) that the latter ‘is one who claims to act and purports to act as a director, although not validly appointed as such.’ By contrast, a shadow director does not so claim or purport. HCL had two titular directors, namely the two Channel Island companies, a fact that Millett J said might itself justify the inference that they were accustomed to act in accordance with the directions of others, in which case those others would be shadow directors. But no such case was pleaded.

The case was argued on the basis (as Millett J was prepared to assume) that sufficient facts were pleaded to show that ET, and perhaps MCP, were each shadow directors of HCL. Taking Dr Hardwick’s case first, it was argued that as he was a director of ET, he was thereby automatically responsible for ET’s conduct in relation to HCL; and was, therefore, himself a shadow director of HCL. Millett J said of that submission, at 184:

The issue before Millett J was whether any case was made out against Dr Hardwick and Mr Thomas that they were shadow directors of HCL. The answer was that it was not. The case made did not go beyond the assertion that they were each directors of ET, which might have been a shadow director of HCL; but it did not follow that ET’s directors were also such shadow directors. In the present case, no assertion was advanced that Mr Holland was a shadow director of the composite companies (unlike section 214, section 212 does not extend to shadow directors); but Mr Knox’s submission was that the principles underlying Millett J’s reasoning apply equally to the assertion that was made, namely that Mr Holland was a de facto director of the composite companies. The answer to that assertion is, he said, that as there was no finding that Mr Holland ever acted in relation to those companies otherwise than in his capacity as a director of their sole corporate director, there was no basis on which he could be held to have been their de facto director. As Millett J put it:

The judge referred briefly (at paragraph [173]) to the Hydrodam case, but did not explain why it did not support Mr Holland’s position. He was apparently more impressed by the guidance given by Evans-Lombe J in Secretary of State for Trade & Industry v. Hall and Nuttall [2006] EWHC 1995 (Ch). That case concerned a claim by the Secretary of State to obtain a disqualification order under the Company Directors Disqualification Act 1986 against Mr Hall and Mr Nuttall as former directors of Mercury Solutions UK Limited (‘Mercury’). Mercury’s only de jure directors were Mr Ahlmann (against whom proceedings were discontinued on undertakings) and Legal Directors Limited (‘LDL’), of which Mr Nuttall was the sole director. Mercury’s business had been managed by Mr Ahlmann and Mr Hall. Mr Hall did not defend the proceedings. Only Mr Nuttall did. His case was that, whilst he was a director of LDL, he had played no part in the business of Mercury. The case against him was that LDL had done nothing to prevent Mercury from committing various defaults (keeping proper books of accounts, making annual returns and so on) and he was said to be liable to disqualification on the basis that he was responsible for LDL’s omissions, he being its sole director and controller. Mr Nuttall challenged the court’s jurisdiction to make a disqualification order against him, because he had never been a director of Mercury.

The question, the judge said at paragraph [16], was ‘whether, in the context of proceedings under [the 1986 Act] it is open to the court to “pierce the corporate veil” and treat Mr Nuttall as a director of Mercury and the 6 companies because, at all material times he controlled LDL which was such a director.’ He rejected the suggestion that Mr Nuttall was a shadow director of Mercury. He referred to and quoted from Hydrodam. He then dealt with the submission that Mr Nuttall was a de facto director of Mercury. The submission was a difficult one because Mr Nuttall had done nothing to hold himself out as such a director. The argument for the Secretary of State was, however, that he should be regarded as such a director, because it was he who controlled LDL – which could and should have performed various duties as a director of Mercury – and his omission to procure LDL to procure the due performance of such duties ought, as a matter of policy, to require him to be regarded as a de facto director of Mercury. Otherwise, it was said, there would be a lacuna in the disqualification legislation: it would prevent the errant directors of a corporate director from being disqualified in consequence of the omissions of the corporate director.

Evans-Lombe J rejected that submission. His reasoning, in paragraph [30], was as follows:

The judge in the present case (in paragraph [175]) cited paragraph 30(iii) of Evans-Lombe J’s judgment and said it led to the clear conclusion that Mr Holland was a de facto director of the composite companies ‘in that he, insofar as he is properly to be regarded as having acted on behalf of Paycheck Directors, clearly caused it to act in such a way as would have caused the latter to be treated as a de facto director were it not already a de jure director.’ He then decided the argument against Mr Holland for the reasons set out in paragraphs [177] and [179], which I have cited. Mr Knox’s submission was that paragraph 30(iii) of the judgment in Nuttall was simply wrong and inconsistent with the principles explained in Hydrodam.

The counter-argument for HMRC, cogently advanced by Mr Green, was that the judge was right in concluding that Mr Holland should be held answerable as a de facto director of the composite companies. It can, Mr Green said, be no answer that Mr Holland’s position was simply that of a director of those companies’ corporate director and that any remedy for wrongdoing by that corporate director must lie against that corporate director alone. That would be fundamentally to undermine the corporate regulatory regime and the public protection which it is intended to achieve. It is, said Mr Green, nonsense to suggest that, in the present case, a remedy is available only against Paycheck Directors, a company with no assets. It was common ground that, had Paycheck Directors not stood between Mr Holland and the composite companies, what Mr Holland actually did would have constituted him a de facto director of those companies. The judge was, it was said, correct to decide that what counted was not the capacity in which Mr Holland did what he did in relation to the composite companies, but whether qualitatively such acts were sufficient to constitute him a de facto director of them. What the court has to do, it was submitted, was to look at what the individual actually did and what he was in a position to prevent. There must, it was said, always be at least one human being responsible for a company’s actions. The approach of Evans-Lombe J in the Hall case was correct, that is to look at the quality of the acts done, not the capacity in which they were purportedly done.

I have come to the conclusion that, with respect, the judge was in error in finding that Mr Holland was a de facto director of the composite companies.

We were referred to statements in the authorities which illustrate the badges of de facto directorship, including the observation by Robert Walker LJ in Re Kaytech International plc, Secretary of State for Trade and Industry v. Kaczer and others [1999] 2 BCLC 351, at 424, that:

I respectfully accept that that is the crucial issue and also that it matters not what the individual is called but what he does. What we were not shown, however, was any authority in which someone who acted as a director of the sole corporate director of a company was, merely by so acting, regarded also as a de facto director of the subject company; and the general statements in the authorities as to the badges of de facto directorship cannot, in my judgment, be treated as deciding that someone so acting will automatically also be so regarded, because they were all focused on different circumstances. The only authorities cited to us that lend any assistance on the question were Hydrodam and Nuttall, although in both the claim of shadow or de facto directorship failed.

The essence of Millett J’s reasoning in Hydrodam is that membership of the board of a corporate director will not, without more, make such member a shadow or de facto director of any company of which the corporate director is a director. Nor in my judgment did Millett J convey or suggest, or intend to convey or suggest, that the requisite more would be satisfied merely by the active participation of the board member in the making of board decisions by the corporate director in relation to the actions of the subject company. On the contrary, I read him as indicating the reverse (see, in particular, the last three quoted paragraphs of his judgment).

I recognise that those paragraphs cannot be said to answer the precise question with which we are faced because Millett J was there focusing on the slightly different question of (in short) whether a director of a corporate shadow director will himself be a shadow director. But in principle it appears to me that the thrust of Millett J’s reasoning applies equally to the question of whether a director of a de jure corporate director will, by acting as such a director, thereby become a de facto director of the subject company. Why, I ask, should such acting have that effect? Accepting that it is constitutionally permissible for a company to have another company as its sole director, I can see no rational basis on which a member of the board of the latter company, who at all times acts in relation to the subject company solely in his capacity as a director of the corporate director, should be regarded as a de facto or shadow director of the subject company. He may well be at the heart of the decision-making in relation to the subject company. But if he is only at such heart in his capacity as the decision-making organ of the corporate director, there is no proper basis for regarding him as a director also of the subject company. He will only become such if he steps outside the confines of his role as a member of the board of the corporate director and acts directly in relation to the affairs of the subject company.

Whilst, therefore, I do not regard Hydrodam as deciding the question of principle with which we are faced, I do regard it as giving a principled steer away from the conclusion that, on the facts of the present case, Mr Holland could properly be regarded as a de facto director of the composite companies. The judge did not apparently so regard Hydrodam and was more influenced by Evans-Lombe J’s decision in Nuttall. I do not, with respect, find the reasoning in Nuttall entirely easy to follow. In paragraph 30(ii) Evans-Lombe J appears to have regarded himself as following Hydrodam although I find it difficult to see in what respect he considered he was doing so; and, for myself, I regard his reasoning as inconsistent with it. Evans-Lombe J’s view, in paragraph 30(ii), was that the board member of the corporate director can become a de facto director of the subject company merely by reason of what he procures the corporate director to do. He appears, therefore, to have been of the view that the mere fact of acting as a director of the corporate director can, or may, result in the member of the corporate director also becoming a director of the subject company. I do not consider that Millett J can be read as suggesting that in Hydrodam and I do not understand why it should be so. The relevant act in relation to the affairs of the subject company is an act directed by the corporate director, not one directed by the latter company’s individual board members. That may be regarded as a distinction of some technicality. But so long as we have a system of company law which recognises the difference between a company and its directors, it is a distinction that must be recognised and respected.

The passage in Evans-Lombe J’s judgment that impressed the judge was paragraph 30(iii), which for convenience I will repeat:

That appears to me, with respect, to establish a somewhat artificial test; and I do not understand its supposed basis. It appears to amount to no more than a proposition that a director of a corporate director will be a de facto director of the subject company if he procures the corporate director to direct the subject company in respects that would make him a de facto director were he to give it such directions directly. In my judgment, the proposition is wrong in principle. I see no reason why a director of a corporate director who is doing no more than discharging his duties as such should thereby become a de facto director of the subject company.

Mr Green emphasised to us that, unless the law is as Evans-Lombe J suggested it is, the regulatory process will be undermined and the protection that such process is intended to confer upon the public will be frustrated. It must, he said, be nonsense that in a case in which the corporate director misdirects the subject company, the remedy can lie only against the corporate director itself, when it may well have no assets to speak of. It must, therefore, be the case that in a case such as the present a remedy will lie against the individual who is a director of the corporate director. He must be regarded as equally answerable before the law for the corporate director’s shortcomings as the corporate director itself.

In principle I have some instinctive sympathy with that submission. But it appears to depend entirely on an assertion as to what the policy of the law should be. It does not, in my view, provide a basis for concluding, contrary to what I regard as the applicable principle, that someone like Mr Holland should be in fact be regarded as a director of a subject company when he has done no more than act as the de jure director of its sole corporate director. The Companies Act 1985 showed itself as well capable of providing, when it wanted to, for circumstances in which an officer of a company is to be as liable as the company itself (see, for example, section 733 in relation to offences by bodies corporate). There is, however, no comparable provision in section 212 of the Insolvency Act 1986 and I am not persuaded that it is for this court to meet that claimed policy need by the artificial expedient of simply treating the directors of a corporate director as de facto directors of the subject company. If there is such a policy need, then it is for the legislature to meet it. It is not for the court to do so.

I might add that not even the Nuttall case provides an example of the application of the underlying policy that Mr Green submitted the court should be applying. If, as he said, it is the policy of the law that an individual should always be answerable for a corporate director’s shortcomings, it is not one that managed to net Mr Nuttall. His company, LDL, had manifestly failed in its duty to ensure that Mercury kept proper books of account but the court rejected the proposition that he, as LDL’s director, should carry the responsibility for that. It rejected it because it felt unable to find that he was a de facto director of Mercury. There is something amiss with Mr Green’s suggested policy if its practical results are as erratic as that. In my judgment, there is no such policy; and it is not for this court to invent one.

I emphasise that nothing that I have said is intended to suggest that there can never be circumstances in which a director of a corporate director can or will so act as to cause himself to be regarded as a de facto director of the subject company. But something more will be required than the mere performance by him of his duties as a de jure director of the corporate director. On the facts accepted by the judge, there was nothing more in the present case.

I would allow Mr Holland’s appeal on the corporate director issue; and would in consequence set aside the judge’s orders made against Mr Holland.

My conclusion on the corporate director issue means that the other grounds of appeal do not arise for decision. Nor does the cross appeal. Because, however, we heard argument on them, I will express my views on them, although more briefly than I would otherwise have done. I do so on the premise, contrary to my conclusion just reached, that Mr Holland was a de facto director of the composite companies.

HMRC’s claim (tempered as I have described) was for the restoration to the composite companies of the dividends unlawfully paid from 24 April 2002 to 19 October 2004. The judge considered that there was a clear distinction between two parts of that period: (a) the period from 24 April 2002 to 19 August 2004 (the day after Mr Holland received Mr Tallon’s advice); and (b) the period from 19 August to 19 October 2004, when the companies went into administration. As for the first period, the judge concluded that ‘either Mr Holland was at no stage liable in respect of the payment of dividends, or if he was, that he ought to be relieved from liability pursuant to Section 727 CA 1985’ (paragraph [236]). There is no cross-appeal against that conclusion.

As for the second period, 19 August to 19 October 2004, the judge took a different view. By then Mr Holland had Mr Tallon’s advice. He knew he had no reasonable prospect of defeating HMRC’s claims for HRCT in respect of the entire trading periods of the companies; he knew he could not justify the continued payment of dividends; and the judge found that the prospect of a deal with HMRC involving the foregoing of the claim to continuing HRCT was no more than speculative. His view was that, from then on, any interim accounts should have made provision for the HRCT that was due (some £3m) and that any dividends paid were unlawful. In paragraphs [261] to [270] he gave his reasons why, as from 23 August onwards, he held Mr Holland to be in principle answerable to HMRC’s claim in respect of the dividends paid during this period and concluded that he ought not to be relieved from liability under section 727 of the Companies Act 1985. On the other hand, he acquitted him from liability in respect of the short period between the receipt of Mr Tallon’s advice on 18 August and 23 August. He explained this (in paragraph [270]) as ‘a few days grace … in order to enable [Mr Holland] to take stock following the Consultation with Mr Tallon QC ….’. The judge did not further explain the basis for this period of grace but I infer he was giving Mr Holland an extension of the section 727 relief granted in respect of the first period.

As for the period from 23 August onwards, the judge did not simply order the payment by Mr Holland to the liquidators of a sum equal to the dividends paid during this period (about £1.5m). He said in paragraph [271] that:

Before coming to the issues, I will set out the material part of section 727 of the Companies Act 1985:

Mr Holland’s second ground of appeal asserts that his causing of the payment of the unlawful dividends as from 19 August 2004 did not subject him to any strict liability to make good to the composite companies the assets so misapplied; it asserts that he would only be answerable to a claim to make such restitution if he had acted negligently in breach of his common law duty of care to the companies or dishonestly in breach of his fiduciary duty to them. The further assertion was that as he paid all the dividends in the honest and reasonable belief that it was in the companies’ interests to do so, he committed a breach of neither duty and so the claim must fail.

Mr Knox accepted that there are authorities indicating that directors are under an unqualified duty not to cause an unlawful and ultra vires payment of a dividend. He referred to In re Exchange Banking Company, Flitcroft’s Case (1882) 21 Ch. D. 591; Re Lands Allotment Company [1894] 1 Ch 616, at 638; In re Sharpe, Masonic and General Life Assurance Company v. Sharpe [1892] 1 Ch 154; Selangor United Rubber Estates Ltd v. Cradock and Others (No 3) [1968] 1 WLR 1555, at 1575; Belmont Finance Corp v. Williams Furniture Ltd and others (No 2) [1980] 1 All ER 393, at 404; Bairstow v. Queens Moat Houses plc and others [2000] 1 BCLC 549, at 555; and Re Loquitur Ltd, Inland Revenue Commissioners v. Richmond and another [2003] 2 BCLC 442, at 471, 472. His submission, however, was that these were all cases in which the defaulting director had either acted dishonestly in causing the payment (and so had been in breach of fiduciary duty), or at least had no reasonable ground for causing it. He submitted also that, in so far as these authorities suggest that the liability is strict, they are inconsistent with another line of authorities to the effect that a director making a misapplication of company assets will only be liable to make restitution if he knew or ought reasonably to have known it was a misapplication. He referred to In re County Marine Insurance Company, Rance’s Case (1870) 6 L.R. Ch. App. 104, at 118, 122; In re Kingston Cotton Mill Company (No 2) [1896] 1 Ch 331, at 345 to 348; Dovey and The Metropolitan Bank (of England and Wales), Limited v. Cory [1901] AC 477; and In re City Equitable Fire Insurance Company, Limited [1925] 1 Ch. 407, at 426

Mr Knox submitted that the latter line of authorities reflects the correct principle. Mr Green, in response, pointed out that the section 727 relief available by the time of this litigation dates only from the Companies Act 1907 (section 32) and so was not available when most of the cases in this line were decided. Now that statutory relief from liability is available in a proper case, he submitted that the court should prefer the view that a director who causes an ultra vires misapplication of company assets should in principle be strictly answerable to a claim to make good the misapplication, subject always to his right to make good, if he can, a claim to relief under section 727.

I regard Mr Green’s submission as a compelling one but find it unnecessary to express a view on its correctness and do not do so. For one thing, we did not have the benefit of full oral argument on the two lines of authorities. But in any event, Mr Knox can only hope to turn his submission to profit if the facts are that Mr Holland acted both honestly and reasonably in causing the unlawful dividends to continue to be paid after 18 August. I do not understand the judge to have made any adverse finding as to Mr Holland’s honesty at any point in the story. But he did reject the assertion that, in so acting, Mr Holland had acted sufficiently reasonably to justify the grant of relief under section 727. In short, Mr Holland failed to surmount the hurdle of reasonableness that he must if he is to be able to invoke the assistance provided by the second line of authority. Mr Knox recognised that and his answer to it was that the judge was wrong to fail to recognise that Mr Holland had acted reasonably at all times; and therefore wrong also to refuse Mr Holland relief under section 727, which is Mr Holland’s fourth ground of appeal. All I need say at this stage is that, in my view, the judge was properly justified in refusing such relief, as I explain below; and I consider it follows that, as this part of Mr Holland’s argument depends on the assertion that Mr Holland acted reasonably, it necessarily fails. I add also, as I will also explain, that I consider the judge was wrong to grant section 727 relief in respect of the ‘short period of grace’ between 18 and 23 August.

In my judgment, therefore, there is nothing in this part of Mr Holland’s second ground of appeal. He was inexcusably at fault in causing the payment of the dividends after 18 August and was in principle liable to make good those misapplications of the companies’ assets.

This is the subject of the second of the two grounds of the cross-appeal. Although it relates only to the period from 19 to 22 August inclusive, it represented the payment of some £140,000 of unlawful dividends. The judge’s justification for the grace period was that Mr Holland was entitled to it in order to take stock of Mr Tallon’s advice.

But Mr Holland did not, in my view, take such stock in any way that justified this head of relief. He simply carried on as before, choosing to take no specialist insolvency advice even though by 19 August it ought to have been obvious that he should do so: that had been made clear by Mr McDonnell, at the consultation with Mr Tallon and by Mr Newton (and see Mr Rees’ letter of 25 August 2004). As (so this part of the discussion assumes) Mr Holland was a director of the composite companies, he had a duty to consider whether he could properly continue to trade them and pay the like dividends as before and his omission to do was irresponsible. If he had ‘taken stock’ of Mr Tallon’s advice in any real sense, he might have done so in one of two ways: (i) by immediately suspending the payment of further dividends and arranging a prompt appointment with an insolvency practitioner as to the proper way ahead; alternatively, (ii) he might have taken the view that he would continue to pay the dividends falling due for payment during those few days (so as not to risk any unnecessary disturbance of the business), whilst also arranging for the earliest possible appointment for advice from an insolvency practitioner. I find it impossible to accept that any insolvency practitioner whose advice had been so sought could have advised the continued payment of dividends and I will assume that, upon such advice, Mr Holland would have stopped their payment. Had the alternative (ii) scenario been the applicable one, I can (just about) see an arguable justification for relieving Mr Holland from liability in respect of dividends paid out during the few days pending the obtaining of advice. But nothing like that happened. All that happened was that Mr Holland put his head into the sand and adopted a ‘business as usual’ policy pending the putting in place of the new structure. In what way, it may be asked, did Mr Holland act reasonably during the grace period so as to deserve the exercise in his favour of the section 727 discretion? The judge did not explain how.

I consider that the judge was in error in giving Mr Holland this grace period: he had not so conducted himself as to deserve it and there was no factual basis for it. Mr Holland may not have behaved dishonestly but he did behave unreasonably. Directors are expected to demonstrate a proper degree of responsibility in conducting the affairs of their companies. Mr Holland’s conduct fell short of the expected standards. I would allow this ground of cross-appeal.

This is the subject of Mr Holland’s fourth ground of appeal and is by way of a complaint that the judge erred in refusing to grant Mr Holland relief against liability for the post-23 August period under section 727. As indicated, I would dismiss this ground of appeal. My reasons substantially mirror what I have said in relation to the ‘days of grace’ cross-appeal.

The point is made, first, that the judge (in paragraph [237]) found Mr Holland to be ‘an honest man placed in difficult circumstances essentially through no fault of his own, ….’. The finding of honesty related, I understand, also to the post-23 August period. The points then made are that: (i) the judge did not expressly take account of Mr Holland’s unchallenged evidence that, if the dividends had been stopped, the composite companies would have been destroyed and that in his view the only sensible thing to do was to continue to trade the companies; (ii) he did not take account of the fact that Mr Holland was throughout advised by, and relying upon the advice, of professionals, none of whom advised him to stop trading or the payment of the dividends – until 13 October, when he stopped both; (iii) that he was wrong to take into account that Mr Holland did not seek specialist insolvency advice, a criticism more properly directed at his professional advisers for not advising it; (iv) that he was wrong, following Mr Tallon’s advice, to regard the prospects of a deal with HMRC as ‘no more than speculative’. In the light of these alleged judicial errors, we are asked to consider afresh the exercise of the section 727 jurisdiction.

In my judgment there is nothing in this criticism of the judge’s judgment. By 9 August Mr Holland had been given the score by Mr McDonnell and Mr Tallon underlined it on 18 August. Both counsel made it plain that Mr Holland should obtain specialist insolvency advice. Mr Newton’s cross-examination showed that he gave him like advice on the train journey home from the consultation with Mr Tallon. As the judge rightly said in paragraph [268.3], ‘the circumstances cried out for specialist insolvency advice.’ Mr Holland cannot rely on any claimed omission of his professional advisers to make this clear to him. On the hypothesis that he was a de facto director of the companies, he should have been aware of the need, once he had had counsel’s advice, to ensure that he could properly continue to authorise the payment of dividends as before. Directors are expected to understand the nature of their duties or at least to understand when circumstances arise which require the obtaining of advice about them. On the face of it, Mr Holland’s conduct in continuing to permit the payment of dividends in circumstances in which there was an accrued liability of some £3m of HRCT to HMRC for which no provision had been made was a serious and inexcusable breach of duty, of which he ought to have been aware and of which any competent insolvency adviser would have advised him. There is also nothing in the point that he thought it sensible to continue to trade the companies. It ought to have been obvious to him that he could not make any sound decision to that effect without first obtaining insolvency advice. Nor is there anything in the point that he entertained the thought of a deal with HMRC: he made no running towards arranging a prompt meeting to discuss such a deal, and it was not until 4 October that he discussed the matter with them. What is obvious is that once Mr McDonnell’s unwelcome news had been confirmed by Mr Tallon, all he had in mind was to trade the companies until the new structure was in place and the business transferred to it. His primary consideration was his interest in preserving a valuable business. The judge was amply justified in concluding that this was not a case for the grant of relief of section 727. I would dismiss this ground of appeal.

This element of the appeal relates to the basis on which the judge ordered compensation in respect of the post-23 August period. It follows from what I have said in relation to the ‘days of grace’ point that I consider that his order necessarily under-compensated the companies. Had he not taken the view he did in relation to the grace period, he would have made a like order in respect of it.

This issue concerns points raised in the balance of Mr Holland’s second ground of appeal and in his third ground; and by HMRC in their first ground of cross-appeal. It is convenient to take the points together. The rival arguments were (by Mr Knox), first, that the only remedy that HMRC were entitled to was one for damages assessed by reference to the net loss sustained by the composite companies by reason of the payment of the unlawful dividends between 19 August and 13 October. The assessment required the increased deficit in HRCT accruing during this period to be set against the savings made by the companies by continuing to trade during this period: those savings were the liabilities that the companies would be likely to have incurred to employee/shareholders and those to whom they provided services if they had stopped trading at the beginning of the period and before the new structure was in place to which all the activity could be switched. Secondly, that the judge had a broad discretion under section 212, both as to liability and quantum, which he should have exercised by refusing HMRC any relief. This was because (a) it was HMRC’s fault that the problems had not been sorted out in 2002, in which event HMRC would not have been entitled to recover any HRCT from the composite companies at all, and (b) if, as it is said they should have done, the companies had stopped trading promptly after 18 August, no dividends would have been paid during the subsequent period and no HRCT would have accrued due during that period. Any award made in favour of HMRC in respect of that period was, therefore, a mere windfall and the judge should have declined to make one.

HMRC’s position, by their first ground of cross-appeal, is that the judge was wrong to exercise his discretion under section 212 by limiting the recovery against Mr Holland in the way he did. He should have ordered him to pay to the companies the full amount of the dividends unlawfully paid out after 19 August. Had these exceeded the amount of the companies’ deficiency (£3.5m), they accept he would have had a discretion under section 212 to limit the order to the amount of that deficiency. But as they did not, he should have ordered payment in the full amount.

Mr Knox’s argument is that as the only basis on which Mr Holland can be criticised for causing the post-18 August dividends to be paid is, so he asserts, that he was in breach of a common law duty of care to the companies, it follows that the correct measure of recovery is damages for that breach, an inquiry as to which will require an investigation into whether the net position of the companies at 19 October (when they went into administration) was worse by reason of the continued trading than it would have been had they stopped trading at, say, 19 August. The point made is that it is said that, had they so stopped trading, they would have been faced with claims from various quarters which were avoided by continuing to trade until 13 October, a period of trading which enabled everyone to be kept sweet whilst the smooth transition into the new structure was arranged. Support for the view that the recovery should be confined to the loss proved to have been caused by the breach of duty was said to be found in the decision of the House of Lords in Target Holdings Ltd v. Redferns (a firm) and Another [1996] 1 AC 421.

The judge was unconvinced by that argument and so am I. Even assuming that the ‘no liability without fault’ line of authority is the guiding one, it does not provide any support for the view that the liability, if proved, of the errant director is merely to compensate the company for the loss assessed on a tortious basis said to have been caused by the misfeasance. It shows, as does the strict liability line, that the basic remedy is one of restitution. That is because directors of a company, if not trustees in the strict sense (because its assets are not vested in them), owe a like duty as a trustee not to misapply the company’s assets and a like duty to make restitution to the company if they do. The principles were summarised, by reference to the authorities, by Etherton J in Re Loquitur Ltd, Inland Revenue Commissioners v. Richmond and another [2003] 2 BCLC 442, at paragraphs [135] to [137] (also a wrongful dividend case).

The judge rejected the submission that the approach in Target Holdings (not a case involving the payment of unlawful dividends) provided any guidance for the purposes of this case and preferred the approach of this court in Bairstow and others v. Queens Moat Houses plc [2001] 2 BCLC 531 (a case that did). In that case the court similarly declined to apply the Target Holdings approach, which it held to be inapplicable to a case in which a director has caused the payment of unlawful, ultra vires dividends. Robert Walker LJ said:

Having considered those passages, the judge said, first, that it had not been ‘suggested that the directors have deliberately or dishonestly paid unlawful dividends.’ There was clearly no suggestion of any dishonesty on the part of Mr Holland in causing their payment (nor was there in relation to the Queens Moat dividends paid in respect of accounts other than the 1991 accounts), but I am not sure I understand in what sense the judge considered that Mr Holland had not deliberately caused their payment. But in any event the judge took the view that:

Having so directed himself that the established remedy in the case before him was an order requiring the reinstatement of the amount of the dividends wrongly paid (some £1.3m), the judge did not make such an order. He simply ordered Mr Holland to pay the amount of HRCT that the companies had not provided for in the period of trading during the period from 23 August (some £144,000). On what basis did he do that?

The answer was, as he said in paragraph [271], that he regarded the core complaint to be the failure to provide for HRCT during this period; and he considered that section 212 conferred upon him a discretion to temper the award available against Mr Holland by so limiting it. He referred to Etherton J’s decision in Re Loquitur (itself drawing on this court’s decision in West Mercia Safetywear Ltd (in liq) v. Dodd and another [1988] BCLC 250) as showing that section 212 confers a discretion on the court in relation to the relief granted. The judge, however, took the view that there was a limit to the extent of the discretion so conferred, even though that limit was not defined in the cases. In explaining why he regarded the discretion as so limited, he said:

If, as the judge there held, there was no discretion under section 212 to grant relief on such a broad basis, why did he relieve Mr Holland against a liability to reinstate the full amount of the unlawfully paid dividends? The explanation was provided in paragraphs [273] and [274], which were added in response to Mr Green’s inquiry upon receipt of the judge’s draft judgment as to whether the judge was implicitly finding that, so long as provision was at least made for the further HRCT falling due, it was lawful for Mr Holland to have caused the payment of dividends without providing for the £3m of unpaid corporation tax already accrued due. The judge’s response to that was no, he was not so finding, and that the dividends paid from 23 August onwards were unlawful and should not have been paid. He continued:

That reasoning was criticised by Mr Green as providing no basis for a departure from the ordinary principle that the remedy against a director whose delinquency was to cause the payment of unlawful and ultra vires dividends is an order for restitution. In my judgment the criticism was, with respect, well-founded. The judge had refused to grant Mr Holland section 727 relief in respect of the dividends unlawfully paid since 23 August, yet made an order that gave him relief against part of the dividends so unlawfully paid. That relief was purportedly granted pursuant to the discretion conferred by section 212, yet the judge had in paragraphs [231] and [232] accepted that the discretion was a limited one that did not enable the grant of relief ‘from liability based on broad and ill-defined ideas of fairness.’ The rationalisation for his decision was (in paragraph [273]) by reference to a consideration of the section 727 relief he would ‘almost certainly’ have granted in a hypothetical scenario which did not in fact happen and which (on Mr Holland’s case) would never have happened and was not considered as a viable option: that was because, unless the employee/shareholders were paid their dividends in full, it was perceived they would promptly have departed the composite companies and switched their allegiance to other groups of composite companies. Mr Green submitted that if, following Mr Tallon’s advice, Mr Holland had been advised by a specialist insolvency adviser that this is the course he should have adopted, and he had done so, there might have been a case for granting the type of section 727 relief to which the judge referred. But this is not what happened.

In my judgment, and with respect to the judge’s reasoning, he fell into error at this point. He had refused Mr Holland section 727 relief in respect of the dividend payments made after 23 August 2004: that refusal was in respect of the entirety of such payments. His justification for nevertheless limiting the section 212 relief to (in effect) merely part of the dividend payments was that in hypothetical circumstances that were never in contemplation he would ‘almost certainly’ have granted section 727 relief in respect of part of the unlawful dividend payments; and it would therefore be within the limited discretion that he regarded section 212 as conferring upon him to limit the section 212 order by reference to the section 727 relief he would probably have granted in those hypothetical circumstances. I respectfully regard that reasoning as flawed and as having led to the wrong result. Having refused section 727 relief in relation to what Mr Holland actually did, there was no basis for, in effect, re-introducing it by reference to a hypothetical course of conduct which did not happen and then purporting to give effect to it by reference to a supposed discretion under section 212. The judge’s order under section 212 should have reflected the wrong that Mr Holland had actually committed – procuring the payment of dividends all of which were ultra vires and unlawful - and the judge’s refusal of section 727 relief in respect of it.

I consider, therefore, that on the face of it the judge’s order for restitution was wrong and that, as asked by ground one of HMRC’s cross-appeal, he ought to have followed the logic of his own findings and directions and ordered Mr Holland to repay the companies the full amount of the dividends unlawfully paid from 19 August 2004 onwards (some £1.5m). That would still have left a deficiency of some £2m in respect of unpaid corporation tax.

Mr Knox disputes that and submitted, in support of Mr Holland’s third ground of appeal, that section 212 gave the judge a much broader discretion than he was prepared to accept that it in fact gave him. The submission was that section 212(3) gives the broadest discretion to make such order ‘as the court thinks just’ and so enabled the court to take a broad view of the merits of the case, to consider questions of causation and the parties’ respective conduct and, in the particular case, to conclude that Mr Holland should not be answerable to the companies for a penny piece.

By way of support for that somewhat sweeping submission, Mr Knox relied on various authorities. First, on In re Exchange Banking Company, Flitcroft’s Case (1882) 21 Ch. D. 519, and in particular what Cotton LJ said at 536. I can, however, detect nothing there that supports the submission, and the particular paragraph relied upon is one with which this court, in Bairstow, had some difficulty (see [2001] 1 BCLC 549, paragraphs [41] to [44]). Second, West Mercia Safetywear Ltd (in liq) v. Dodd and another [1988] BCLC 250. In that case, WM Ltd, in liquidation, was a subsidiary of D Ltd. D was a director of both companies. Shortly before D put both companies into liquidation, WM Ltd’s bank account was in credit whereas D Ltd’s account was overdrawn, the liability being guaranteed by D. D procured the transfer of £4,000 from WM Ltd’s account to D Ltd’s account, the purpose being to reduce his own potential liability under his guarantee. The liquidator of WM Ltd applied under the predecessor of 212 (section 333 of the Companies Act 1948) for an order against D for the reinstatement of the £4,000. This court, reversing the judge, held that D’s procuring of the payment of the £4,000 was a breach of duty and that ‘therefore the declaration sought in the notice of motion ought to be made against’ D (per Dillon LJ at 253c). Dillon LJ then said:

The third case to which Mr Knox referred was Re Westlowe Storage and Distribution Ltd [2000] BCC 85 in which, at 870, Hart J said that ‘the remedy under s. 212 is discretionary’ but made orders against Mr Lowe in respect of the matters where he had held him to be in breach of duty. I derive no support from that case for Mr Knox’s submission as to the breadth of the section 212 discretion. Fourth, Mr Knox relied on In re Continental Assurance Co of London plc [2001] 2 BPIR 733, at paragraph 393, in which Park J said that section 212:

Mr Knox cited no authority, however, for the proposition that the section 212 discretion is so wide as to extend to the grant of relief against liability as opposed to a discretion as to the appropriate order to make once liability has been established. In West Mercia Dillon LJ indicated no more than that there was a discretion ‘over the matter of relief’, which may well also be all that Park J was saying. A case illustrating that this is the limit of the discretion is Etherton J’s decision in Re Loquitur. That too was a case in which the respondents were held liable in a claim under section 212 for causing the payment of an unlawful dividend but, at [2003] 2 BCLC 442, paragraph [245], the judge explained the discretion in terms showing that he regarded it as extending merely to the nature of the relief to be granted, a view he regarded as supported by Dillon LJ’s guidance in West Mercia. He went on to explain why the justice of the case before him did not require him to order the respondents to repay the full amount of the dividend.

In my judgment Mr Knox’s submission as to the breadth of the discretion conferred by section 212 is wrong; and that the judge was correct to reject it. As Chadwick LJ explained in Cohen and another v. Selby and another [2001] 1 BCLC 176, paragraph [20]:

Just as section 212 does not create new rights and obligations, nor in my judgment does it create new discretionary defences as to liability. It certainly makes no reference to any. The essence of the discretion that section 212(3) undoubtedly does confer is, as Dillon LJ said in West Mercia, ‘over the matter of relief’, by which he meant in respect of the order that the court may make once liability has been established; and he gave brief illustrations as to why the court might, in this respect, conclude that ‘the wind should be tempered’ on the particular facts of any case. Mr Green rightly accepts that the court has such a discretion. Taking the present case, suppose that HMRC had made good their claim that Mr Holland had wrongfully caused the payment of unlawful dividends during the whole period since 24 April 2002 and Mr Holland had been refused section 727 relief. The dividends so paid amounted to some £13m, but it was no part of HMRC’s case that Mr Holland might be ordered to repay such an amount to the companies. To order him to do so would produce an apparent injustice: such an order, if satisfied, would not merely meet the deficiency of £3.5m in the companies, it would (costs apart) leave nearly £10m in the companies for distribution to their members, thereby in effect paying them the same dividends a second time. That could readily be said to be unfair, and the court would obviously have a discretion to confine any order to the amount of the deficiency, or £3.5m (a discretion that might, however, not have been available had the claim been brought by the composite companies themselves: see Bairstow [2001] 2 BCLC 531, at paragraphs [44] to [48], per Robert Walker LJ).

It follows in my view that the submission that the judge should have had regard to broad discretionary matters in deciding whether to find Mr Holland liable as claimed by HMRC is mistaken. The judge’s task was to decide whether the cause of action on which it was based was established; if it was, whether Mr Holland was entitled to be relieved against liability to any extent under section 727; and, subject to that, to decide the fair order to make under section 212. That is what he did, although I have given my reasons for concluding that he misdirected himself at the last stage.

In my judgment, the judge ought to have ordered Mr Holland to repay the full amount of the dividends unlawfully paid during the period 19 August to 19 October 2004, amounting to some £1.3m. That would not have resulted in any ‘windfall’ for the companies’ shareholders. They would not have seen a penny of it. It would merely have gone towards reducing the overall deficiency of some £3.5m represented by the companies’ liabilities to HMRC for HRCT. Nor would it have represented any ‘windfall’ for HMRC either. The tax became due because the trading by each of the companies generated profits that were chargeable to such tax. The tax that became due is tax that was and is payable and Mr Holland’s arguments that it only became due because the tax structure of the companies was flawed and/or because he did not stop them trading earlier are without substance. It is not HMRC’s fault that the tax avoidance scheme was flawed or that he chose to continue trading the companies for as long as he did. It was not HMRC’s fault that Mr Holland chose to cause the continuing payment of unlawful, ultra vires dividends after 18 August. Those payments, to the employee/shareholders of the companies, should never have been made.

Were I not disposed to allow Mr Holland’s appeal on the ‘corporate director’ ground, I would have dismissed his appeal on all grounds and allowed HMRC’s cross-appeal on both grounds. As, however, I consider that Mr Holland is entitled to succeed on his ‘corporate director’ ground, it follows that he is the overall winner. I would allow his appeal, set aside those parts of the judge’s order that were adverse to him and dismiss the cross-appeal.

I am grateful for the careful analysis of the facts in the judgment of Rimer LJ, and I will not repeat them. I agree with his conclusions and his reasoning with respect to all aspects of the case save two. The first is the relatively trivial issue whether the judge was entitled to give a few days’ grace to the appellant. The second, and more important point, is whether the judge was entitled to apply section 212 of the Insolvency Act 1986 in the way that he did so as to limit the amount which Mr Holland had to repay to the company.

On the central issue, namely whether Mr Holland was a de facto director, I entirely agree with the analysis of Rimer LJ. I do not see how the directors of a corporate director, acting purely in that capacity, can by virtue of that fact alone be constituted de facto directors of the company of which the corporation is the sole director.

It is pertinent to note that in In re Bulawayo [1907] 2 Ch. 458 the case in which Warrington J accepted that a corporate body could be a director, it was urged upon the court that there would be insufficient accountability if this were permitted. Much of the argument of Mr Green, counsel for the Revenue, appeared to me, with respect, to be making essentially the same point save that since it is now firmly established that corporate bodies can be directors, he sought to remedy the alleged defects by pinning individual liability on the persons who are directors of the corporate director. That will no doubt be justified where it is successfully asserted that the corporate director is a sham company; but no such allegation was made here. (Section 155 of the Companies Act 2006 now requires that there be at least one natural person as a director, but that was not in force at the material time.)

Moreover, as Millett J, as he then was, pointed out in In re Hydrodam (Corby) Limited [1994] 2 BCLC 180, a de facto director is someone who purports to act as a director although not validly appointed as such. I do not see how that description could possibly apply to Mr. Holland. He has not sought to act as a director of any of the composite companies or purported to be, or represented himself as, a director of any of them. If anything, he would be a shadow director i.e. exercising influence in fact but concealing that influence behind the cloak of the corporate director of the composite companies, Paycheck Directors. However, the decision in Hydrodam, which I respectfully consider to be wholly consistent with principle, is inconsistent with him being so treated merely because he effectively determines the policy of Paycheck Directors by virtue of his position on the board of that company. Accordingly, in my view no liability attaches to Mr. Holland for the unlawful payment of dividends.

Strictly the other points do not arise, but I respectfully agree with Rimer LJ that they need to be considered in case we are wrong on the first point, and we heard full argument on them.

There is an unusual feature of this case which has moulded the arguments before us. Usually when dividends are paid to shareholders, nothing is secured in return. Accordingly, the unlawful payment of dividends from undistributed profits will generally deplete the company’s assets by the amount distributed.

In this case, however, the dividends are in substance remuneration for services rendered by the shareholder employees. So the actual loss resulting from the unlawful payment of dividends is reduced by the extent to which profits are generated by the company’s activities. This has enabled Mr Knox QC, counsel for Mr Holland, to run various arguments which it seems to me are really premised on the basis that we should treat these payments of dividend as something paid out in the ordinary course of business notwithstanding their formal status as dividends.

Mr Knox contends that there are three grounds on which, even assuming that Mr Holland was a de facto director, he should not have been found liable for misfeasance in effecting the distribution of the dividends. First, he contended that since the payments were in the best interests of the composite companies, since they would otherwise have had to cease trading with significant potential contractual liabilities, they should not be treated as unlawful at all. That argument is wholly unsustainable if the payments are properly to be treated as the payments of dividends, as they undoubtedly are. It is ultra vires for the company to make payments to shareholders out of undistributable profits; it is forbidden both at common law and under statute (see section 263 of the Companies Act 1985) and since it is illegal it cannot be done even with the consent of all the members: see e.g. Aveling Barford Ltd v Perion Ltd.[1989] BCLC 626. So the fact that the directors may consider the payment to be in the company’s best interests cannot conceivably be a defence, at least if the officer appreciates that the payment, properly characterised, is the return of undistributed profits by way of dividend.

A second and related point is an argument that at common law there is no personal liability if a director acts honestly and reasonably. Rimer LJ has identified the nature of the competing arguments in paragraphs 82-83 of his judgment. As he points out, this submission in any event adds nothing, at least in this case, to the protection afforded to Mr Holland because he would be entitled to relief from liability under section 727 if he had acted honestly and reasonably and ought fairly to be excused.

This in turn raises Mr Knox’s third submission, which is that the judge erred in concluding that section 727 was not applicable to the final period of trading after August 2004. The essence of his case appears to be that Mr Holland received no unequivocal professional advice telling him to cease trading. I entirely agree with Rimer LJ that the judge was manifestly entitled to conclude that this section was inapplicable in the circumstances and that Mr Holland had not acted reasonably with respect to this final period of trading.

Mr Holland also submitted that even if there were liability, he should not be required to pay any compensation by way of relief, for two reasons. The first is that the company suffered no loss as a result of the unlawful payment of dividends. As I have said, normally there would be no benefit resulting from such payments, and the loss would be clear and unarguable, but in this case there is a benefit because the payment of dividends keeps the companies going. This both generated profits and, so it is argued, limited the risk of potential liabilities for breach of contract claims from employees and other third parties which would have arisen had trading suddenly ceased. That has let in the argument that the wrongdoers should only be liable for the resulting loss to the company, after having taken into account the profits and other potential savings which would have been foregone had no dividends been paid. The second reason is a related one, namely that in the exercise of discretion under section 212, the court ought to have refused to grant any relief. I deal with this point below (para 135) when considering other arguments with respect to the application of that subsection.

As to the first argument, I agree with Rimer LJ (see paras.93-98) that the principle of loss is not the basis on which a fiduciary has to account for the misuse of corporate funds. The basic obligation placed on someone who misapplies company funds is to restore the moneys wrongly paid out, although for reasons I give below, I think that it is permissible to have regard to the question of loss when considering the issue of relief under section 212.

I now turn to the two areas where I have the misfortune to disagree with Rimer LJ. They concern the two issues which are the subject of the cross-appeal by the Revenue, namely the fact that the judge allowed a few days’ grace, and that he did not require Mr Holland to account for the full amount of the dividends unlawfully paid out after 23 August, which amounted to some £1.3 million, but only the relatively small sum which constitutes the tax that ought to have been paid in the period of trading post 23 August 2004, namely £144,000.

The judge held that even once the advice had been given by Mr Tallon QC on 18 August confirming that the Revenue were right in contending that there was a tax liability, it was not unreasonable for Mr Holland to take a few days to consider the implications of that advice. Rimer LJ says that this is not legitimate because it is plain that Mr Holland had simply buried his head in the sand and had never given thought to doing anything other than to continue running the businesses until a new corporate structure, eliminating the companies’ tax liabilities, had been put in place. Mr Holland’s conduct was unreasonable at all stages once he knew what the true tax position was.

It seems to me that the relevant question is from when Mr Holland should be taken to have been acting unreasonably so that the defence of section 727 was no longer applicable. If the facts as relied upon by Rimer LJ were the incontrovertible facts, I would agree with his conclusion. However, in my view there was evidence justifying the judge’s analysis. He recounted the evidence of Mr Newton about the latter’s conversation with Mr Holland on the train immediately after the meeting with Mr Tallon. Mr Newton said that he did not tell Mr Holland to cease trading; indeed, he said that the risk of paying unlawful dividends had to be balanced against the risk of breach of contract claims if the businesses were suddenly to cease. These were, as Mr Newton observed, matters for Mr Holland to think about. The judge himself observed that these were unpalatable options for Mr Holland who found himself in “very difficult circumstances”, and I think the judge was entitled to infer, as I believe he did, that they must have been occupying Mr Holland’s mind (even although he did not recall the particular conversation with Mr Newton). The fact that Mr Holland subsequently determined on an unreasonable course of action, even if that was by default because he was unwilling properly to address the issue, does not mean that he was acting unreasonably before adopting that position. In my view it was realistic for the judge to recognise that some short period to assess the position was not unreasonable and that in fact some assessment of the options must have been made notwithstanding that it did not cause Mr Holland to take active steps to remedy the wrongdoing.

That section is set out by Rimer LJ in paragraph 44 above. It confers a discretion on the court to require the wrongdoer to make a payment of such amount as seems to the court to be just. I accept that this does not confer an unprincipled discretion, or one that allows the judge to fix the amount by reference to vague notions of fairness. But the concept of justice is a broad one; I agree with the observations of Etherton J (as he then was) in Re Loquitur Ltd, Inland Revenue Commissioners v. Richmond and another [2003] 2 BCLC 442 para 245, that section 212 confers a discretion with respect to relief, in addition to the defence to liability in section 727, enabling the court to require the wrongdoer to restore “what is just in all the circumstances.”

Rimer LJ has set out the judge’s reasoning on this issue but it is important and so I repeat it here:

“[273] … However, I ask myself what the result would have been if the Composite Companies had, following the meeting with Mr Tallon QC, continued to pay dividends but, contrary to the practice prior thereto, made full provision for ongoing HRCT by reducing the dividend payments accordingly. The dividends would, of course, still have been unlawful because the outstanding substantial liability for HRCT as at 23 August 2004 would, for the purposes of Section 270 CA 1985, and interim accounts prepared for the purposes of Section 273 CA 1985, have had to have been provided for so as to comply therewith. However, having found as I have in relation to the period prior to 23 August 2004, and given the particular circumstances of the present case and the connection between the payment of dividends and work actually done by the contractors/shareholders, and charged for by the Composite Companies on an ongoing basis as described above, generating what would otherwise have been distributable profits, I would almost certainly, in the circumstances, have been prepared to relieve Mr Holland from liability under Section 727 CA 1985.

[274] On the basis of my findings, the culpable failure in the present case is, as I have said, the continued payment of dividends after 23 August 2004 without providing for the further HRCT that became due as a result of the continued trading, with the consequence that HMRC lost out to the extent thereof. To limit the award against Mr Holland in the way that I propose to do in accordance with my discretion under Section 212 1A 1986 is consistent with my finding as to the scope of the discretion under Section 212 referred to in paragraph [231] above, and, I believe, the approach taken by Etherton J in Re Loquitur (supra), and so limiting the award reflects the circumstances in which the present claims have been brought, namely by HMRC in a liquidation in which they are the only substantial creditor…….”

Rimer LJ points out, and I agree, that the analysis in paragraph 273 is not satisfactory. The judge is there positing the possibility that following 23 August, had the companies continued paying dividends after making the relevant tax payments to the Revenue with respect to the ongoing tax liabilities, he would have been minded to relieve Mr Holland from liability with respect to that liability under section 727. This was so notwithstanding that strictly the payments would still have involved the unlawful distribution of undistributable profits because of the substantial outstanding debt owed to the Revenue as a consequence of the payment of dividends in the earlier periods. However, the judge had already found that it was not reasonable to continue to pay these dividends once it became clear that to do so was unlawful. It is difficult to see on what basis he could have concluded that it would have been reasonable to make what were on any view, and to the knowledge of Mr Holland, unlawful dividend payments, even if the ongoing tax liabilities were met.

However, I think that paragraph 274 is identifying a distinct and justifiable basis for the judge’s conclusion. The judge is there identifying the culpable conduct as the payment of dividends after 23 August. Until then, Mr Holland was relieved from liability by section 727. Thereafter, what the Revenue has lost from his unlawful conduct is the tax liability referable to that conduct. Had the businesses been closed on the 23 August, the Revenue would have got nothing.

In the context of this case, and given the unusual feature that these dividend payments generated profits for the companies, the payment of the dividends to employees did not deprive the Revenue of sums that would otherwise have been available to meet the composite companies’ tax liabilities. In those circumstances, and given that the Revenue was the only creditor, the judge concluded that it was not just that Mr Holland should have to restore all the dividends paid out so as to make good earlier liabilities incurred in the period in respect of which the judge found he had a defence.

In my judgment it was open to the judge to conclude that limiting the payments in this way achieved what was just in all the circumstances. He plainly thought it would be harsh to require Mr Holland to have to restore all the unlawful dividends so as to meet the liabilities to the Revenue even with respect to the period when he had been acting reasonably, and I personally have some sympathy with that view. In any event, it seems to me that it was open to the judge to exercise his discretion in the way he did. I accept that this involved recognising the fact that unusually the payment of these dividends benefited the company perhaps as much as the recipients and did not simply dissipate corporate funds which would otherwise have been available to meet the tax liabilities; but that is surely an important feature of the case when all the circumstances have to be considered. Moreover, there was no evidence that creditors other than the Revenue were adversely affected in any way.

Mr Knox made the bold submission that the judge ought, in the proper exercise of the section 212 discretion, to have concluded that Mr Holland should have been relieved from the obligation to pay anything at all. Even if it could ever be justified for a judge to find that it was just that no compensation should be paid for a breach of duty - and it is difficult to envisage such circumstances - this was not such a case. Mr Holland had as by then he well knew continued to act unlawfully by paying the dividends, and he had not even secured the payment of the relevant tax incurred during that period. The submission that the only proper exercise of discretion under section 212 was for the judge to relieve him of the obligation to make any payment, including by reference to those tax liabilities, seems to me to be quite hopeless.

For the reasons I have given, I would allow the appeal on the grounds that Mr Holland was not a de facto director. On all other matters, however, I would uphold the analysis of the judge below and I pay tribute to his careful and thoughtful judgment.

I have read in draft the judgments of Rimer and Elias L.JJ. For the reasons they have given, I agree that Mr Holland was not a de facto director. I would allow his appeal accordingly. It is, therefore, strictly speaking unnecessary to consider the other interesting but difficult points that arise on the appeal but since my Lords have expressed their views, I offer my tentative conclusions secure in the knowledge that they can be disregarded henceforth as obiter dicta.

I would allow Mr Holland these few days grace to take stock of his position and to reflect on whether he had to close down the business to the considerable inconvenience of the companies’ employees and those who relied upon their services and at risk of claims for damages being brought by them. Mr Holland’s anger with his advisers was made manifest on the return train journey after the devastating consultation with Mr Tallon Q.C. He was, not surprisingly, in a state of some shock and, not having been given any advice other than that he should seek specialist insolvency advice, he was, in my view, reasonably entitled to take a few days to gather his thoughts, in the reasonable expectation that his local advisers upon whom he had been so reliant, would, as a matter of urgency, tell him what to do. The fact that he thereafter unreasonably buried his head in the sand should not deny him the benefit of the first few days grace.

The view I take of this issue is this. HMRC’s claims were brought under section 212 of the Insolvency Act 1986 which provides a “summary remedy against delinquent directors” as follows:

“212(1) This section applies if in the course of the winding up of a company it appears that a person who –

(a)

is or has been an officer of a company [and for the purposes of this part of the judgment Mr Holland is assumed to be an officer], …

has misapplied any money … of the company …

(3)

The court may on the application of … any creditor … compel him –

(a)

to repay … any part of [the money] …”

On the other hand section 727 of the Companies Act 1985 gives the court “power to grant relief in certain cases”. It provides:

“(1)

If in any proceedings for negligence, default, breach of duty or breach of trust against an officer of the company … it appears to the court hearing the case that that officer … is or may be liable in respect of the negligence, default, breach of duty or breach of trust, but that he has acted honestly and reasonably, and that having regard to all the circumstances of the case … he ought fairly to be excused for the negligence, default, breach of duty or breach of trust, that court may relieve him, either wholly or partly, from his liability on such terms as it thinks fit …”

The key question for me is how those two sections interrelate. It seems to me that section 212 provides the remedy where the officer of the company has misapplied any money but then gives the court a discretion to compel payment of part of the money only if that is what the court thinks just. It is a twofold exercise: (1) establish how much has been misapplied, in this case arguably the whole amount of the dividends which should not have been paid, assessed at £1.3M, and (2) decide according to the justice of the case whether only a part of it should be repaid. The criterion for the exercise of the latter discretion is the justice of the case depending on all the circumstances of the case. I have no difficulty in considering justice to be a proper, principled guide to the exercise of discretion.

Having fixed the amount of the liability under section 212, it seems to me that the court has a further dispensatory power afforded by section 727 provided the officer has acted both honestly and reasonably. I am content to proceed upon a basis that, the few days of grace apart, Mr Holland acted unreasonably in continuing to pay these dividends and so he is not entitled to any further mercy under that provision. But it still leaves open the question of how much it is just that he should pay under section 212.

One can paint justice on a broad canvas. Justice seems to me to require in this case that what Her Majesty’s Revenue and Customs should recover from this summary remedy, which gives the Exchequer a right of action as a creditor, is no more than the amount of tax the Revenue have actually lost. The Official Receiver or the liquidator, who stand in the shoes of the company, may have had a different claim and I can see why they might possibly recover the whole of the payments improperly made. On one view of it, if Mr Holland had done what, according to the tax laws, he ought to have done, then he would have shut down the business immediately and the Revenue would not have received anything. But he traded on. His erroneous decision has thus benefited the Revenue and provided a fortuitous windfall. But in my view HMRC should receive no more than would have been coming into their coffers, namely the amount of the Higher Rate Corporation Tax. It does not seem to me to be just that they should receive the whole of the dividend payments. That seems to me to be an enrichment which it would be unjust for them to receive.