In the matter of the Lehman Companies (No. 2)

before

Lord Neuberger, President
Lord Mance
Lord Clarke
Lord Sumption
Lord Toulson

JUDGMENT GIVEN ON

24 July 2013

Heard on 14, 15 and 16 May 2013

Appellant William Trower QC Tom Smith Andrew Mold (Instructed by Herbert Smith Freehills LLP)	Respondent Raquel Agnello QC Jonathan Hilliard Thomas Robinson (Instructed by The Pensions Regulator)
	Respondent Richard Sheldon QC Michael Tennet QC Felicity Toube QC (Instructed by Hogan Lovells International LLP)
Appellant Mark Phillips QC Stephen Robins James Walmsley (Instructed by Linklaters LLP)	Respondent Raquel Agnello QC Jonathan Hilliard Thomas Robinson (Instructed by The Pensions Regulator)
Appellant Robin Dicker QC Paul Newman QC Daniel Bayfield (Instructed by Linklaters LLP)	Respondent Raquel Agnello QC Jonathan Hilliard Thomas Robinson (Instructed by The Pensions Regulator)
Respondent Gabriel Moss QC Nicolas Stallworthy QC David Allison (Instructed by Travers Smith LLP)	Respondent Barry Isaacs QC (Instructed by Weil, Gotshal & Manges)

LORD NEUBERGER (with whom Lord Mance, Lord Clarke and Lord Toulson agree)

Introductory

1. These two appeals raise questions of some significance arising out of the interrelationship of the statutory schemes relating to the protection of employees’ pensions and to corporate insolvency.

2. The background to the two appeals is, in very summary terms, as follows:

1. Many UK registered members of the Lehman group of companies, and all the UK registered members of the Nortel group of companies, have gone into insolvent administration;
2. (a) One of those Lehman group companies entered into service contracts with, and ran a pension scheme for the benefit of, employees who worked for other group members;

(b) The Nortel group included a company which had a pension scheme, and which was insufficiently resourced to fund that scheme;

3. The pension scheme (“the Scheme”) in each case was a final salary scheme, which appears to be, and to have been for some time, in substantial deficit;
4. The Pensions Regulator subsequently initiated machinery under the Pensions Act 2004 to require certain other group members (“the Target companies”) to provide financial support for the Scheme;
5. That machinery has been held up so it can be decided whether the liability under such a requirement would rank (a) as an expense of the Target companies’ administrations, (b) pari passu with the Target companies’ other unsecured creditors, or (c) as neither;
6. Under option (a) the liability would rank ahead of the unsecured creditors, and may well be paid in full; under option (b) it would rank equally with those creditors; under option (c) it would rank behind them, and would probably be worthless;
7. Briggs J and the Court of Appeal (in a judgment given by Lloyd LJ) concluded that option (b) was not open to them, and preferred option
   1. to option (c);
8. The issue now comes before the Supreme Court.

3. This judgment starts by explaining the relevant statutory provisions relating to pensions, which are mostly in the Pensions Act 2004 (“the 2004 Act”), in a

description largely based on the exposition in the judgment of Briggs J, [2010] EWHC 3010 (Ch), [2011] Bus LR 766, paras 7-41. It then deals with the statutory provisions and rules relating to insolvency, in the Insolvency Act 1986 (“the 1986 Act”) and the Insolvency Rules 1986 (SI 1986/1925) (“the Insolvency Rules”), largely drawing on what Lloyd LJ said in the Court of Appeal, [2011] EWCA Civ 1124, [2012] Bus LR 818, paras 20-23 and 39. Next, it will explain the facts, in a summary reflecting what Briggs J said at paras 47-54 of his judgment.

4. After a short discussion, the judgment will then turn to consider whether the liabilities in the present cases would rank pari passu with the unsecured creditors of the Target companies. It will then consider whether those liabilities rank as expenses of the administration. Finally, it will address the power of the court under the 1986 Act and the Insolvency Rules to vary the priority of the liabilities.

The relevant statutory provisions relating to pensions

5. In order to protect employees from the adverse consequences of an under- funded occupational pension scheme, (i) the Social Security Act 1990 introduced a statutory debt regime by amending the Social Security and Pensions Act 1975, and

(ii) the Pensions Act 1995 (“the 1995 Act”) introduced a minimum funding requirement regime. These were perceived to be inadequate in some respects, and the 2004 Act introduced a financial support direction (“FSD”) regime. The regimes under these Acts were introduced against the backdrop of European Directives, which require member states to take measures to protect the interests of employees or ex-employees in relation to pension rights in the event of their employer’s insolvency.

Section 75 of the 1995 Act

6. Although it is the FSD regime under the 2004 Act which is of central importance on these appeals, section 75 of the 1995 Act is highly relevant. It provides that upon the happening of various events, which include an “insolvency event”, an amount equivalent to any shortfall in the assets of an occupational pension scheme (a “scheme”) as against its liabilities, which exists immediately prior to the relevant event, is to be a debt, known as a “section 75 debt”, due from the employer to the trustees of the scheme (the “trustees”). Under the section as originally drafted, an “insolvency event” was limited to the employer going into insolvent liquidation, but the 2004 Act extended the expression to include going into administration. In this judgment I shall similarly use the expression to cover going into administration or going into insolvent liquidation.
7. Section 75(8) provides that a section 75 debt is not to be regarded as a preferential debt for the purposes of the 1986 Act. Section 75(4A) states that a section 75 debt is to be taken, for the purposes of an employer’s insolvency, to arise immediately before the occurrence of the insolvency event.

The 2004 Act: the Regulator and the PPF

8. The 2004 Act introduced both the Pensions Regulator (“the Regulator”) and the Pension Protection Fund (“the PPF”).

9. The Regulator is a body corporate established by section 1, and, by section 4, it is given wide regulatory functions. When exercising any of those functions, the Regulator is required by section 100 to “have regard to”:

“2(a) the interests of the generality of the members of the scheme to which the exercise of the function relates, and

(b) the interests of such persons as appear to the Regulator to be directly affected by the exercise.”

Section 5(1) defines the Regulator’s main objectives, which include protecting the benefits of members of schemes, and reducing the risk of compensation having to be paid by the PPF.

10. The PPF is financed from levies upon schemes. It operates by assuming the assets and liabilities of a deficient scheme, and then paying its members compensation at a prescribed rate (generally less than the full rate promised under the relevant scheme), using the industry-wide levies for the purposes of meeting the shortfall between the deficient scheme’s assets and the prescribed level of compensation.

The 2004 Act: the FSD regime and FSDs

11. It was perceived that the creation of the PPF might encourage some employers to arrange their affairs so as to throw the burden of pension scheme deficiencies upon the PPF, which would unfairly burden other schemes by increasing the amount of the levies. An example of such an arrangement is where a group of companies uses a single company (a “service company”) to employ people who then work for other group companies. In such a case, the employees’

pension rights could be regarded as unfairly prejudiced if, by comparison with the resources of other group companies, the service company had very limited resources to meet a section 75 debt.

12. The FSD regime was designed to mitigate such problems. In a nutshell, it enables the Regulator in specified circumstances (i) to impose, by the issue of a FSD to some or all of the other group companies (known as “targets”), an obligation to provide reasonable financial support to the under-funded scheme of the service company or insufficiently resourced employer, and (ii) to deal with non-compliance with that obligation by imposing, through a Contribution Notice (a “CN”), a specific monetary liability payable by a target to the trustees.

13. The detailed provisions of the FSD regime are contained in sections 43 to 51 of the 2004 Act, and in the Pensions Regulator (Financial Support Directions etc) Regulations 2005 (SI 2005/2188) (“the FSD Regulations”).

14. Section 43 is of central importance. Subsection (1) explains that the FSD regime extends to all occupational pension schemes other than money purchase schemes and certain other prescribed schemes. Section 43(2) contains the so-called “employer condition”, and provides as follows:

“The Regulator may issue a [FSD] … in relation to such a scheme if the Regulator is of the opinion that the employer in relation to the scheme—

1. is a service company, or

2. is insufficiently resourced,

at a time determined by the Regulator which falls within subsection

(9) (‘the relevant time’).”

Section 43(9) and the FSD Regulations define “the relevant time” as any time within a period of two years before the date of the determination of the Regulator to issue the FSD in question. It is known as “the look-back date”.

15. “Service company” is defined in section 44(2) as being a company within a group of companies which, by reference to its turnover, can be seen to be

principally engaged in providing the services of its employees to other member companies in the group.

16. Section 44(3) to (5) and the FSD Regulations explain that an employer is “insufficiently resourced” if two tests are satisfied. The first is that the value of its resources is less than 50% of the estimated section 75 debt in relation to a scheme, the amount of the shortfall being “the relevant deficit”. Secondly (limiting the situation to those involving companies), there must be a company which has (or two or more companies which between them have) resources not less than the relevant deficit, and which is (or are), inter alia, a company which is (or companies which are) “connected with, or an associate of” the employer (section 43(6)(c)).

17. The 2004 Act and the FSD Regulations contain detailed provisions as to the manner in which a person’s resources are to be assessed. Whereas the resources of an employer are incapable of being defined as having a negative value, the resources of persons associated or connected with the employer may be so defined. The formula for determining whether the insufficiently resourced condition is satisfied is known as the “rich man/poor man test”.

18. The “employer condition” operates entirely by reference to the look-back date chosen by the Regulator, rather than at the time when the FSD is issued (“the issue date”). Accordingly, the fact that, as at the date the FSD is issued, an employer may have ceased to be a service company, or the rich man/poor man test is not met, would not preclude a FSD.

19. As to the “target”, section 43(4) provides that a FSD in relation to a scheme may be issued to one or more persons, but subsection (5)(a) limits the issue of a FSD to persons falling within subsection (6) at the relevant time (i.e. the look-back date). Section 43(6)(a) and (c) respectively limit that class to the employer itself and, for present purposes, to “a person … who is connected with or an associate of the employer” at the look-back date. It is therefore irrelevant that, by the issue date, one or more targets which had the requisite net worth to satisfy the rich man part of the rich man/poor man test as at the look-back date may no longer be solvent. Further, section 43(5)(a) does not limit the range of potential targets to those which satisfy the rich man part of the rich man / poor man test at the look- back date.

20. Section 43(5)(b) states that a FSD can only be issued to a particular target if the Regulator is “of the opinion that it is reasonable to impose the requirements of the direction on that person”; this is often called the “reasonableness condition”. Section 43(3) states that a FSD should:

“[require] the person or persons to whom it is issued to secure –

1. that financial support for the scheme is put in place within the period specified in the direction,

2. that thereafter that financial support or other financial support remains in place while the scheme is in existence, and

3. that the Regulator is notified in writing of prescribed events in respect of the financial support as soon as reasonably practicable after the event occurs.”

“Prescribed events” include an insolvency event affecting the employer and any target, and any failure to comply with the requirements of the FSD.

21. Section 45(1) and (2) define “financial support” as “one or more” of the following arrangements:

“(a) an arrangement whereby … all the members of the group are jointly and severally liable for the whole or part of the employer’s pension liabilities in relation to the scheme;

2. [a legally binding] arrangement whereby … a company … which meets [certain] requirements and is the holding company of the group is liable for the whole or part of the employer’s pension liabilities in relation to the scheme;

3. an arrangement which meets [certain] requirements and whereby additional financial resources are provided to the scheme …;

4. such other arrangements as may be prescribed.”

Subsection 45(3) states that the Regulator may only approve any arrangement if satisfied that it is “reasonable in the circumstances”. Subsection 45(4) explains that “the employer’s pension liabilities” include, but are not limited to, the employer’s section 75 debt.

22. A FSD will not itself either contain or be accompanied by a specification of what would constitute reasonable arrangements. It will simply require that the target secures that financial support for the scheme is put in place. It is for the target (alone or in conjunction with other targets) to propose reasonable arrangements for written approval by the Regulator. What the FSD must specify, pursuant to subsection 43(3)(a), is the period within which financial support for the scheme is to be put in place. By contrast, the period during which that support is to remain in place is, by reference to subsections 43(3)(b) and (10), the whole of the period until the scheme is wound up.

23. Section 43(7) provides that, when deciding “whether it is reasonable to impose the requirements” of a FSD, the Regulator is to “have regard to such matters as the Regulator considers relevant including, where relevant, the following matters”:

“(a) the relationship which the person has or has had with the employer (including … whether the person has or has had control of the employer …),

2. in the case of a person falling within [section 43(6)(c)], the value of any benefits received … by that person from the employer,

3. any connection or involvement which the person has or has had with the scheme,

4. the financial circumstances of the person,

…. .”

24. The FSD regime is capable of applying to almost any company within a group which has a service company, or a potentially insufficiently resourced employing company, with a potentially under-funded scheme. Accordingly, the consequential contingent liabilities it creates could undermine the financial stability of potential targets. Section 46(2) attempts to mitigate this problem by providing for applications to the Regulator to determine that:

“(a) the employer in relation to the scheme would not be a service company for the purposes of section 43,

2. the employer in relation to the scheme would not be insufficiently resourced for the purposes of that section, or

3. it would not be reasonable to impose the requirements of a financial support direction, in relation to the scheme, on the applicant.”

Once issued, such a clearance statement binds the Regulator in relation to the power to issue a FSD unless there has been a relevant change of circumstances from those described in the application.

The 2004 Act: the FSD regime and CNs

25. The Regulator can issue a CN where there has been non-compliance with a FSD. Whereas a single FSD can be issued in relation to a scheme (albeit to one or more targets), CNs are only to be issued on a target by target basis. Thus, section 47(4)(d) expressly contemplates that a CN may be issued to one target, where others have proposed arrangements in response to a FSD which have received the Regulator’s approval.

26. Section 47(3) imposes a reasonableness condition upon the issue of a CN to a particular target. Potentially relevant considerations are listed in subsection (4). In addition to those listed in section 43(7) in relation to FSDs, there are two further considerations, namely:

“(a) whether the person has taken reasonable steps to secure compliance with the financial support direction [and] …

4. the relationship which the person has or has had with the parties to any arrangements put in place in accordance with the direction (including, where any of those parties is a company within the meaning of subsection (11) of section 435 of the Insolvency Act 1986, whether the person has or has had control of that company within the meaning of subsection (10) of that section).”

27. By contrast with a FSD, a CN is required to be specific as to the amount payable by the target. By section 47(2) the notice must state that the target is under a liability to pay the scheme trustees or managers a specified sum. By section 48, that sum is to be either the whole or a specified part of the amount which the Regulator estimates to be the amount of the section 75 debt at the time of non-

compliance with the FSD section 75 debt (if by then crystallised) or (if not crystallised) the Regulator’s estimate of what such debt would be if it had crystallised at that date.

28. Section 49(3) provides that “[t]he sum specified in the notice is to be treated as a debt due from the person to the trustees or managers of the scheme.” Provision is also made for the Regulator (or, in specified circumstances, the Board of the PPF) to exercise any powers of the trustees or managers to recover the debt.

29. CNs can be issued to two or more targets, and to create joint and several liability for a specified amount. Section 50 enables the Regulator to restrain the trustees or managers of the scheme from pursuing recovery of the section 75 debt while, at the same time, a CN is being enforced. Section 50(6) ensures that any payments under a CN are treated as reducing the amount of the section 75 debt. Finally, section 50(9) enables the Regulator to reduce the amount specified in a CN where, for example, there have in the meantime been payments of part of the section 75 debt.

The 2004 Act: Procedure

30. The 2004 Act and the FSD Regulations lay down a fairly elaborate procedural code for the implementation of functions of the Regulator, including the FSD regime. The functions of the Regulator are divided between regulatory functions, which are exercisable by its executive arm, and reserved regulatory functions, which must be exercised by its Determinations Panel (“the DP”). Decisions to issue a FSD and a CN are reserved functions, whereas the decision whether to give written approval to proposed arrangements under section 45 is not.

31. Although the Regulator has a degree of discretion as to its procedure, in relation to the FSD regime it must comply with what is called in section 96 the “standard procedure”, which involves, as a minimum:

“2(a) the giving of notice to such persons as it appears to the Regulator would be directly affected by the regulatory action under consideration (a ‘warning notice’),

2. those persons to have an opportunity to make representations,
3. the consideration of any such representations and the determination whether to take the regulatory action under consideration,

4. the giving of notice of the determination to such persons as appear to the Regulator to be directly affected by it (a ‘determination notice’),

5. the determination notice to contain details of the right of referral to the Tribunal … .”

32. The issue of a FSD and a CN must each be subject to this procedure. The Tribunal is now the Upper Tribunal (Tax and Chancery Chamber), from which an appeal lies to the Court of Appeal. By section 103(4) the Tribunal must, on a reference, “determine what (if any) is the appropriate action for the Regulator to take in relation to the matter referred to it”. The Tribunal must therefore approach the issue before it afresh rather than by way of reviewing the decision of the Regulator or the DP.

33. It is worth briefly summarising the timescale involved in these procedures. Before implementing the standard procedure, the Regulator must identify a pension fund which appears to be at risk, and investigate whether the conditions for the implementation of the FSD regime are satisfied; it must then address all matters relevant to the exercise, including the reasonableness condition for the issue of a FSD to each potential target. The ensuing standard procedure potentially involves six stages at which the target and others can make representations, namely (i) after a warning notice, (ii) following a determination (before the Tribunal), (iii) following a FSD, (iv) after a warning notice that a CN may be issued, (v) upon a determination that it should be issued (before the Tribunal), (vi) even after the issue of a CN, an adjustment may be asked for in the light of payments by others. At every stage, the Regulator or the Tribunal is required to have regard to the interests of the target as a person directly affected.

The Insolvency legislation

Administration and liquidation

34. For present purposes, there are two relevant types of corporate insolvency procedure, administration and liquidation. Liquidation, or winding up, has always been a feature of company law, and it can be invoked whether or not a company is insolvent, although insolvent liquidations are more common. Administration was

first introduced by the 1986 Act. At that time, it did not allow for distributions to creditors of the company within the administration. If the administration did not succeed in rescuing the company, it was expected that a winding-up would follow, and the available assets would be distributed to creditors within the liquidation. The Enterprise Act 2002 (“the 2002 Act”) rendered it possible for assets to be distributed to creditors by administrators, so that a winding-up can be avoided. (Conversely, a company which is in liquidation may now come out of it and go into administration.) There may be companies which go into liquidation without having been in administration, but most of those companies with which the present cases are concerned (apart from one which is not insolvent) are in administration, and may or may not eventually go into liquidation (and two of the Lehman group companies are neither in administration nor in liquidation).

35. In general, the unsecured debts of a company after an insolvency event are payable pari passu to the relevant creditors, who claim payment by proving for their debts. There has to be a cut-off date to determine the class of creditors who are to participate in the distribution of the company’s available net assets. As the law stood as regards the companies with which these appeals are concerned, the cut-off date for claims in a liquidation is the date on which the company goes into liquidation, whether or not the liquidation was immediately preceded by an administration. The cut-off date for claims in an administration is the date on which the company entered administration. Under this regime, if an administration is followed immediately by a liquidation, the debts provable in the liquidation would include any which arise during the administration, although debts provable in the administration would be limited to those arising before the administration.

36. Before turning to the relevant statutory provisions, two points may be worth noting in passing.

37. First, the position described in para 35 above has now changed. The cut-off date for claims in a liquidation, which follows an administration started after 5 April 2010, is the date when the administration began. The same issue as arises in these appeals can still arise. However, there will no longer be an artificial distinction between the positions where the company proceeds from administration to winding-up and where it does not. The change will tend to increase the importance of the dispute as to the correct treatment for insolvency purposes of the liabilities arising under a FSD or a CN.

38. Secondly, in relation to the companies in the present cases, it is common ground that if a liability of such a company arises during the administration, and a winding-up were to follow later, that liability can be the subject of proof in the liquidation.

The relevant provisions of the 1986 Act and the Insolvency Rules

39. In a liquidation of a company and in an administration (where there is no question of trying to save the company or its business), the effect of insolvency legislation (currently the 1986 Act and the Insolvency Rules, and, in particular, sections 107, 115, 143, 175, 176ZA, and 189 of, and paras 65 and 99 of Schedule B1 to, the 1986 Act, and rules 2.67, 2.88, 4.181 and 4.218 of the Insolvency Rules), as interpreted and extended by the courts, is that the order of priority for payment out of the company’s assets is, in summary terms, as follows:

1. Fixed charge creditors;
2. Expenses of the insolvency proceedings;
3. Preferential creditors;
4. Floating charge creditors;
5. Unsecured provable debts;
6. Statutory interest;
7. Non-provable liabilities; and
8. Shareholders.

40. So far as expenses of an insolvency are concerned, rule 12.2 of the Insolvency Rules (“rule 12.2”) states that:

“(1) All fees, costs, charges and other expenses incurred in the course of winding up, administration or bankruptcy proceedings are to be regarded as expenses of the winding up or the administration or, as the case may be, of the bankruptcy.”

41. As to expenses in a liquidation, rule 4.218 (“rule 4.218”) provides:

“(1) All fees, costs, charges and other expenses incurred in the course of the liquidation are to be regarded as expenses of the liquidation. …

3. [T]he expenses are payable in the following order of priority –

1. expenses … properly chargeable or incurred by the official receiver or the liquidator in preserving, realising or getting in any of the assets of the company or otherwise in the preparation or conduct of any legal proceedings … or in the preparation or conduct of any negotiations; …

(e) the cost of any security provided by a … liquidator; …

13. any necessary disbursements by the liquidator in the course of his administration …;

14. the remuneration or emoluments of any person who has been employed by the liquidator to perform any services for the company

…;

15. the remuneration of the liquidator …;

16. the amount of any corporation tax on chargeable gains accruing on the realisation of any asset of the company; …

(r) any other expenses properly chargeable by the liquidator in carrying out his functions in the liquidation.”

42. The equivalent provision in relation to the expenses of an administration is rule 2.67(1) (“rule 2.67(1)”), which states that “The expenses of the administration are payable in the following order of priority”, namely,

“(a) expenses properly incurred by the administrator in performing his functions in the administration of the company;

(b) the cost of any security provided by the administrator in accordance with the Act or the Rules; …

(d) any amount payable to a person employed … to assist in the preparation of a statement of affairs ….; …

6. any necessary disbursements by the administrator in the course of the administration …;

7. the remuneration or emoluments of any person who has been employed by the administrator to perform any services for the company ….;
8. the remuneration of the administrator …;

(j) the amount of any corporation tax on chargeable gains accruing on the realisation of any asset of the company …. .”

Where the assets of the company are insufficient to meet the totality of the expenses, rule 2.67(2) and (3) gives the court power to “make an order as to the payment out of the assets of the expenses incurred in the administration in such order of priority as the court thinks just”.

43. Turning to unsecured debts and liabilities which are not expenses or preferential debts, rules 4.181 and rule 2.69 of the Insolvency Rules apply to liquidations and administrations respectively, and make it clear that, in so far as they are provable, they “rank equally” and, if there is insufficient money to meet them all, that they are to be “abate[d] in equal proportions among themselves”.

44. In relation to what constitutes a provable debt, rule 12.3 of the Insolvency Rules in its form which applies to the instant administrations (“rule 12.3”) is headed “Provable debts”, and it provides as follows:

“(1) Subject as follows, in administration, winding up and bankruptcy, all claims by creditors are provable as debts against the company …, whether they are present or future, certain or contingent, ascertained or sounding only in damages. …

(3) Nothing in this Rule prejudices any enactment or rule of law under which a particular kind of debt is not provable, whether on grounds of public policy or otherwise.”

45. Rule 13.12 of the Insolvency Rules (“rule 13.12”) is of critical importance on these appeals and it states:

“(1) ‘Debt’ in relation to the winding up of a company, means… any of the following –

1. any debt or liability to which the company is subject … at the date on which the company went into liquidation;
2. any debt or liability to which the company may become subject after that date by reason of any obligation incurred before that date;

…

2. For the purposes of any provision of the Act or the Rules about winding up, any liability in tort is a debt provable in the winding up, if either –

1. the cause of action has accrued … at the date on which the company went into liquidation; or

2. all the elements necessary to establish the cause of action exist at that date except for actionable damage.

3. For the purposes of references in any provision of the Act or the Rules about winding up to a debt or liability, it is immaterial whether the debt or liability is present or future, whether it is certain or contingent, or whether its amount is fixed or liquidated, or is capable of being ascertained by fixed rules or as a matter of opinion … .

4. … except in so far as the context otherwise requires, ‘liability’ means (subject to paragraph (3) above) a liability to pay money or money’s worth, including any liability under an enactment, any liability for breach of trust, any liability in contract, tort or bailment, and any liability arising out of an obligation to make restitution.

5. This Rule shall apply where a company is in administration and shall be read as … if references to winding-up were references to administration.”

The relevant facts

The Lehman group

46. The Lehman group was a very substantial international financial concern which notoriously collapsed on 15 September 2008. The main London-based group companies in the group were placed into administration that day, and I will refer to the administrators as “the Lehman Administrators”. The ultimate parent company of the Lehman group is Lehman Brothers Holdings Inc. (“LBHI”), a

company incorporated in Delaware USA, which commenced Chapter 11 bankruptcy proceedings in September 2008, and emerged from them in March 2013. The main UK operating company is Lehman Brothers International (Europe) (“LBIE”), an unlimited company. The principal Lehman employer company within the UK, providing employees on secondment for most of the group’s European activities, based in London, is Lehman Brothers Limited (“LBL”).

47. When LBL went into administration on 15 September 2008, it crystallised a section 75 debt in relation to the Lehman Brothers Pension Scheme of approximately £120m. LBL is a shareholder in LBIE, and therefore liable without limit for LBIE’s liabilities. Both LBIE and Lehman Brothers Europe Limited (“LBEL”), the other main London operating company, are subsidiaries of Lehman Brothers Holdings plc (“LBH”), which is itself wholly owned by Lehman Brothers UK Holdings Limited (“LBUKH”), which is in turn an indirect subsidiary of LBHI.

48. Shortly after the Lehman group crash, the Regulator began investigations, with the consent of the Lehman Administrators, into the Lehman companies pursuant to notices under section 72 of the 2004 Act. Warning notices were issued to a number of Lehman group companies on or after 24 May 2010 on the ground that LBL was a “service company”, and the other requirements of section 43 of the 2004 Act were satisfied. There was then an oral hearing in September 2010 before the DP (at which the Lehman Administrators’ solicitors attended to observe, but made no submissions). A determination was then made by the DP on 13 September 2010 that a FSD should be issued against six Target companies, namely LBHI, LBIE, LBEL, LBH, LBUKH and Lehman Brothers Asset Management (Europe) Limited, which is no longer part of the Lehman group and is now called Neuberger Newman Europe Limited. The FSD process in relation to the Lehman companies is now stayed until after the outcome of these applications.

The Nortel Group

49. Prior to its collapse in January 2009, the Nortel group carried on a very substantial international telecommunications, computer network and software business. Its ultimate parent company is Nortel Networks Corporation (“NNC”) based in Canada. Its main Canadian operating company was Nortel Networks Limited (“NNL”) and its substantial USA business was headed by Nortel Networks Inc. (“NNI”), a direct subsidiary of NNL.

50. The group’s principal operating company in the UK was Nortel Networks UK Limited (“NNUK”) which is also a direct subsidiary of NNL. Since June 2000 it is principal Nortel employer in relation to the Nortel Networks UK Pension Plan

(“the Nortel Scheme”). NNUK had a number of subsidiaries incorporated in various European countries. In addition, the European business was also carried on by certain European subsidiaries of NNL, including the applicants Nortel Networks SA, Nortel Networks France SAS and Nortel Networks (Ireland) Limited. At the time of the group’s collapse in January 2009, NNUK’s section 75 debt crystallised in an amount of about £2.1bn.

51. Upon the group’s collapse, NNC and NNL sought protection under Canadian bankruptcy law to facilitate the reorganisation of the group for the benefit of its creditors. On the same day NNI was placed into Chapter 11 bankruptcy in the United States, whilst NNUK, fifteen of its subsidiaries and the three European subsidiaries of NNL referred to above were placed into administration in England.

52. The English administrators of the nineteen Nortel companies (“the Nortel Administrators”) have cooperated with other Nortel group office-holders worldwide, in the process of selling the Nortel group’s businesses along business rather than corporate demarcation lines and total global realisations of approximately US$7.5bn have been made.

53. The Regulator’s investigations into the Nortel Scheme began in early 2009, with the benefit of information provided by the Nortel Administrators. A warning notice was issued on 11 January 2010 to twenty-nine Target companies in the Nortel group, on the basis that NNUK was “insufficiently resourced” and the other requirements of section 43 of the 2004 Act were satisfied. Representatives of the Nortel Administrators were present as observers, but did not make any representations at the oral hearing before the DP on 2 June 2010. After that hearing, the DP issued a determination notice on 25 June 2010 deciding that a FSD should be issued to the applicant Nortel companies, together with certain other Target companies. Following a reference to the Tribunal by the applicant Nortel companies and certain other of the Target companies, the automatic stay of the FSD process means that no FSD has yet been issued to those referring Target companies. The Tribunal proceedings have been informally stayed pending the outcome of these applications.

Overview

54. The issue in both appeals is how the administrators of a target should treat the target’s potential liability under the FSD regime (and in due course the liability under a CN) in a case where the FSD is not issued until after the target has gone into administration. The courts below both held that the potential liability constituted an expense of the administration, falling within category (2) as

described in para 39 above, so that it took priority over the normal run of unsecured creditors – and even over the preferential creditors. Four possibilities have been canvassed before us. The first is that the courts below were right. The second is that the potential liability is an ordinary provable unsecured debt, ranking pari passu with other unsecured debts falling within category (5). The third possibility is that it is not a provable debt within rule 13.12, and therefore it falls within category (7). The fourth possibility is that, if the third is correct, then the court could and should direct the administrators to treat the potential FSD liability more favourably.

55. Counsel representing the various parties very sensibly divided up the issues between them, so as to ensure that there was no repetition, and it is right to record the court’s gratitude for the way the appeals were argued.

56. Both Briggs J and the Court of Appeal felt constrained by a consistent line of authority, of which the most recent is R (Steele) v Birmingham City Council [2006] 1 WLR 2380, from holding that the potential liability as a result of a FSD issued after the commencement of an insolvent administration or liquidation (which I will refer to as “an insolvency event”) could constitute a “provable debt” within rule 12.3, although it appears that they would have so held if they had felt able to do so (see eg Briggs J’s “reluctance” at para 191 of his judgment). They also considered that the effect of the House of Lords decision in In re Toshoku Finance UK plc [2002] 1 WLR 671 was that the potential liability was to be treated as an expense of the administration.

57. Before this court, it was common ground that the potential liability under a FSD could not be both a provable debt and an expense of the administration, but there was discussion as to which should be considered first. In some cases, a liability which would otherwise be a provable debt can be, on special facts, an expense of the administration or liquidation (as in In re ABC Coupler and Engineering Co Ltd (No 3) [1970] 1 WLR 702), which may seem to suggest that the expense issue should be considered first. However, in the light of the common ground in this case, it appears to me that it is appropriate to consider the provable debt issue first, although it would be wrong not to address the expense question as well.

58. Before I turn to examine in detail the arguments on the two issues, it is right to say that, at any rate on the face of it, the sensible and fair answer would appear to be that the potential liability of a target, under a FSD issued after an insolvency event, and in particular the liability under a CN issued thereafter, should be treated as a provable debt. There seems no particular sense in the rights of the pension scheme trustees to receive a sum which the legislature considers they should be

entitled to receive having any greater or any lesser priority than the rights of any other unsecured creditor.

59. It is common ground that if a CN had been issued in respect of a company before an insolvency event, it would give rise to a provable debt, and the courts below considered that, if a CN were issued after an insolvency event, it would give rise to a provable debt if it was based on a FSD issued before the insolvency event. It appears somewhat arbitrary that the characterisation and treatment of the liability under the FSD regime should turn on when the FSD or CN happens to have been issued, if it is based on a state of affairs which existed before the insolvency event.

60. The notion that the potential liability under the FSD regime should be a provable debt if the FSD is issued after the administration or liquidation is supported by the fact any section 75 debt would itself be a provable debt, and not a preferential debt, in any insolvent liquidation or administration of an employer. That is clear from the provisions summarised in para 7 above. It would be strange if the employer company’s statutory obligation to make good a shortfall in its employees’ pension scheme ranked lower in its insolvency than the more indirect statutory obligation of a target to make that deficiency good ranked in the target’s insolvency. Indeed, it would be somewhat surprising if there was any significant difference in the treatment of the two types of obligation, in the light of the interrelationship between the FSD regime and the section 75 debt as evidenced for instance by section 50(6) of the 2004 Act.

61. If the decisions below were correct, it would also mean that the legislature had given the Regulator a significantly valuable and somewhat arbitrary power, in what may (in the light of what is said in para 63 below) be an admittedly rare case in practice. Where the Regulator is proposing to issue a FSD in respect of a company not yet in administration or liquidation, it would be well advised to wait for the insolvency event, if the decisions below are right, because the amount recoverable under a subsequent CN would inevitably be greater than under a CN issued following a FSD issued before the insolvency event.

62. The liability under the FSD regime could be said to be some sort of indirect liability for past wages of employees, as pensions are often treated as deferred pay. However, quite apart from the fact that that argument involves a considerable stretch (not least because the liability is not that of the employer or former employer), it would prove too much. If the potential liability under consideration in these appeals counted as expenses, they would rank ahead of past wages and holiday pay, which have preferential status – ie they would fall within category (3), not (5), in para 39 above.
63. It also seems unlikely that it can have been intended that liability under the FSD regime could rank behind provable debts. One would have expected that FSDs and CNs would normally be issued in respect of insolvent companies (that certainly appears to have been the invariable experience so far); accordingly, it would mean that, save in very unusual cases, nothing would be paid in respect of most FSDs issued after an insolvency event. Further, it would be a relatively unusual case where a FSD, let alone a CN, was issued before an insolvency event. As Briggs J said, the investigations, steps and opportunity for representations, required by the 2004 Act, as summarised in para 34 above, are such that a FSD would rarely be issued for many months after the Regulator is first aware of a possible problem, and the time before a CN could be issued could easily be much more than a year.

64. With those preliminary observations, I turn to consider whether the liability under a FSD issued after a target has gone into administration or liquidation (ie after an insolvency event) is (i) a provable debt, or (ii) an expense of the administration/ liquidation, and (iii) if it is neither, whether the court can require the administrator or liquidator to treat it as if it was.

Is the liability under a FSD issued after an administration a provable debt?

65. In the light of the reference in rule 13.12(4) to “any liability under an enactment”, it appears clear that a liability under a CN, which on any view imposes a duty on the target to pay a sum of money, would be capable of giving rise to a liability. The question, therefore, is whether the potential liability under a FSD which was issued after an insolvency event is capable of being a “liability” falling within the ambit of rule 13.12(1).

66. The definition of “provable debt” in rule 12.3 is strikingly wide, particularly when the rule is read together with rule 13.12, which defines “debt”. It is therefore unsurprisingly not in issue that the consequences for a target of, or at least following, the issuing of a FSD constitute a “liability” within the meaning of rule

13.12. However, where a FSD is issued after the target goes into administration or liquidation, the argument which has to be addressed is whether any potential liability thereby created falls within rule 13.12(1), because, if it does not, then it cannot constitute a “debt” for the purposes of the rule, and therefore cannot be a provable debt for the purposes of rule 12.3.

67. The primary argument advanced in support of the contention that the potential liability under a FSD notice issued after the insolvency event gives rise to a provable debt was based on the contention that it would be “a … liability to which the company may become subject after [the insolvency event] by reason of

any obligation incurred before that date” within rule 13.12(1)(b). However, it was also argued that, in the light of the very wide ambit of rule 13.12(3) and its reference to “future”, “contingent” and “matter of opinion”, the potential liability would fall within rule 13.12(1)(a), as the potential liability was, as at the date of the insolvency event, a “liability to which the company is subject”.

Does the potential liability fall within rule 13.12(1)(a)?

68. It is convenient first to deal with the argument that potential liability under a FSD issued against a company after the start of its administration or liquidation gives rise to a liability which falls under rule 13.12(1)(a). This argument is based on the simple proposition that the risk of being issued with a FSD is a contingent liability, and is therefore a “liability” for the purposes of rule 13.12(1)(a) as a result of rule 13.12(3). If this argument was right, it would avoid the possible problem thrown up by the closing ten words of rule 13.12(1)(b).

69. The argument would be easy to understand were it not for rule 13.12(1)(b). Para (b) of rule 13.12(1) contains a limitation, in that it provides that, if a company in liquidation or administration becomes subject to a liability after the date of the insolvency event, then that liability can only be treated as a “debt” under that paragraph if it arises “by reason of any obligation incurred before that date”. If para (a) of rule 13.12(1) could apply to a liability which arises after the insolvency event, then it would not only render para (b) otiose, but it would also effectively override this limitation. In other words, the very limitation which rule 13.12(1)(a) is being invoked to avoid represents the reason why rule 13.12(1)(a) cannot be invoked.

70. It is fair to say that it is somewhat ironic to invoke para (b) to limit the ambit of para (a), when it would appear that the purpose of para (b) is to extend the ambit of para (a). However, the provisions of the Insolvency Rules, and of each rule of those Rules, have to be read in a sensible and coherent way, and one has to read paras (a) and (b) so that they work together. I agree with the view expressed by David Richards J in In re T & N Ltd [2006] 1 WLR 1728, para 115, that para

(a) is concerned with liabilities to which the company “is subject” at the date of the insolvency event, whereas para (b) is directed to those liabilities to which it “may become subject” subsequent to that date, and that there is no overlap between these two categories.

71. Accordingly, if there is a “debt or liability” in this case, it cannot fall within para (a): the issue is whether it falls within para (b).

Does the potential liability fall within Rule 13.12(1)(b)?

72. There is no doubt that the liability which is imposed on a target on the issuing of a FSD after the commencement of its administration or liquidation is a liability for the purposes of rule 13.12(1)(b), as it is a “liability under an enactment” within rule 13.12(4). The question is, however, whether it can be said to be a liability which arose “by reason of any obligation incurred before” the insolvency event.

73. In these cases, a Target company’s liability under the FSD scheme arises because it was a member of a group of companies, which, to put it very loosely, fell within the scope of the regime (as the group included a company which had a pension scheme, and that company was a service company, or insufficiently resourced). In order for the liability in issue to fall within rule 13.12(1)(b), therefore, the fact that the Target company was a member of such a group must amount to a sufficient “obligation incurred” within the meaning of that rule, before the target went into administration. Timing is no problem in the present cases, because each of the Target companies in the Lehman and Nortel groups were members of a group which fell within the scope of the regime, as I have used that expression, well before they went into administration.

74. That issue thus centres on the meaning of the word “obligation” in rule 13.12(1)(b). The meaning of the word “obligation” will, of course, depend on its context. However, perhaps more than many words, “obligation” can have a number of different meanings or nuances. In many contexts, it has the same meaning as “liability”, but it clearly cannot have such a meaning here. Indeed, in the context of rule 13.12, it must imply a more inchoate, or imprecise, meaning than “liability”, as the liability is what can be proved for, whereas the obligation is the anterior source of that liability.

75. Where a liability arises after the insolvency event as a result of a contract entered into by a company, there is no real problem. The contract, in so far as it imposes any actual or contingent liabilities on the company, can fairly be said to impose the incurred obligation. Accordingly, in such a case the question whether the liability falls within para (b) will depend on whether the contract was entered into before or after the insolvency event.

76. Where the liability arises other than under a contract, the position is not necessarily so straightforward. There can be no doubt but that an arrangement other than a contractual one can give rise to an “obligation” for the purposes of para (b). That seems to follow from rule 13.12(4). As Lord Hoffmann said, (albeit in a slightly different context) in relation to contingent liabilities arising on a

liquidation, in Secretary of State for Trade and Industry v Frid [2004] 2 AC 506, para 19, “[h]ow those debts arose – whether by contract, statute or tort, voluntarily or by compulsion – is not material”.

77. However, the mere fact that a company could become under a liability pursuant to a provision in a statute which was in force before the insolvency event, cannot mean that, where the liability arises after the insolvency event, it falls within rule 13.12(1)(b). It would be dangerous to try and suggest a universally applicable formula, given the many different statutory and other liabilities and obligations which could exist. However, I would suggest that, at least normally, in order for a company to have incurred a relevant “obligation” under rule 13.12(1)(b), it must have taken, or been subjected to, some step or combination of steps which (a) had some legal effect (such as putting it under some legal duty or into some legal relationship), and which (b) resulted in it being vulnerable to the specific liability in question, such that there would be a real prospect of that liability being incurred. If these two requirements are satisfied, it is also, I think, relevant to consider (c) whether it would be consistent with the regime under which the liability is imposed to conclude that the step or combination of steps gave rise to an obligation under rule 13.12(1)(b).

78. When deciding whether a particular state of affairs or relationship is sufficient to amount to the “incur[ring]” of an “obligation”, “by reason of which” the liability arose, considerable assistance can, I think, be gained from the majority decision in Winter v Inland Revenue Commissioners, In re Sutherland (dec’d) [1963] AC 235. That case was concerned with whether an arrangement was within the expression “contingent liabilities” in section 50 of the Finance Act 1940. As Lord Reid explained at p 247, at the relevant date, “the position of the company … was that, by applying for and accepting allowances in respect of these ships, it had become bound by the statute to pay tax under a balancing charge when it ceased to use these ships in its trade, if the moneys which it received for them exceeded any expenditure on them which was still unallowed”.

79. In those circumstances, the majority concluded that the obligation was a contingent liability as at the relevant date. Lord Reid said this at p 248:

“[I]f an Act says I must pay tax if I trade and make a profit, I am not before I begin trading under a contingent liability to pay tax in the event of my starting trading. In neither case have I committed myself to anything. But if I agree by contract to accept allowances on the footing that I will pay a sum if I later sell something above a certain price I have committed myself and I come under a contingent liability to pay in that event.”

80. Reference to a passage in Lord Hodson’s dissenting speech highlights the effect of this reasoning. At p 257, he said that he thought that “the risk of attracting liability is not enough and the argument involves a misconception of what is meant by ‘contingent liabilities’ in their context”, and went on to point out that “[t]here may be no day of reckoning; the ships may never be sold; if there is a sale there may be a balancing allowance not a balancing charge”. This contrast is also highlighted by what Lord Guest (who agreed with Lord Reid) said at p 264:

“The claim for initial allowances for what has been described as depreciation is the voluntary choice of the taxpayer, but, once he has obtained such allowances, he is automatically involved by the operation of law in the payment of balancing charges, if the assets are parted with at a price greater than the written down value in the circumstances defined in section 292 of the Income Tax Act, 1952”

81. It is true that in Sutherland, the House of Lords was concerned with the meaning of “contingent liabilities” in the context of estate duty, whereas these appeals are concerned with the meaning of “obligation” from which a contingent liability derives in insolvency legislation. It was suggested that the reasoning of Lord Reid should not, therefore, be relied on here. I do not agree. Lord Reid gave a characteristically illuminating and authoritative analysis of an issue of principle. It appears to me that the issue of (i) what is a contingent liability and (ii) what is an obligation by reason of which a contingent liability arises, are closely related. In Sutherland the House had to decide whether what a company had done was sufficient, in Lord Reid’s words, to have “committed [it]self” to a contingent liability. As I see it, that is much the same thing as having incurred an obligation from which a contingent liability may arise, for the purposes of rule 13.12(1)(b).

82. I note that the approach to contingent liabilities adopted in Sutherland was considered helpful in two cases concerned with insolvency law decided by judges experienced in the field– Pennycuick J In re SBA Properties Ltd [1967] 1 WLR 799, 802D-803E, and David Richards J in In re T & N Ltd, [2006] 1 WLR 1728, paras 48-61. In the latter case, the judge pointed out at para 61 that the case before him was, as these cases are, “in one important respect a stronger case” than Sutherland, because “the majority did not regard as decisive that the liability to pay the balancing charges would arise only as a result of the company’s own choice to sell the ships. In this case there is no question of volition.”

83. The reasoning of Lord Reid, and of Lord Guest, in Sutherland self- evidently supports the argument that the potential FSD regime liabilities in the present cases fall within rule 13.12(1)(b), even where the FSD is not issued until after the relevant insolvency event. More specifically, if one asks whether those

potential liabilities of the Target companies in these two appeals satisfy the requirements suggested in para 77 above, it appears to me that the answer is yes.

84. As to the first requirement, on the date they went into administration, each of the Target companies had become a member of a group of companies, and had been such a member for the whole of the preceding two years – the crucial look- back period under the 2004 Act. Membership of a group of companies is undoubtedly a significant relationship in terms of law: it carries with it many legal rights and obligations in revenue, company and common law.

85. As to the second requirement, by the date they went into administration, the group concerned included either a service company with a pension scheme, or an insufficiently resourced company with a pension scheme, and that had been the position for more than two years. Accordingly, the Target companies were precisely the type of entities who were intended to be rendered liable under the FSD regime. Given that the group in each case was in very serious financial difficulties at the time the Target companies went into administration, this point is particularly telling. In other words, the Target companies were not in the sunlight, free of the FSD regime, but were well inside the penumbra of the regime, even though they were not in the full shadow of the receipt of a FSD, let alone in the darkness of the receipt of a CN.

86. So far as the third requirement is concerned, I would simply refer back to the points made in paras 58-63 above.

The earlier authorities

87. I should refer to the authorities which the Court of Appeal and Briggs J understandably held bound them to reach a contrary conclusion. Those authorities were mostly concerned with individual bankruptcy rather than corporate insolvency. However, the meaning of the expression “debt” in the two regimes is very similar: rule 12.3 applies to both, and section 382 of the 1986 Act has a very similar definition of provable debt for bankruptcies as rule 13.12 has for liquidations.

88. In a number of cases, it has been held that, where an order for costs was made against a person after an insolvency process had been instituted against him, his liability for costs did not arise from an obligation which had arisen before issue of the bankruptcy proceedings, even though the costs order was made in proceedings which had been started before that insolvency process had begun – see for instance In re Bluck, Ex p Bluck (1887) 57 LT 419, In re British Gold

Fields of West Africa [1899] 2 Ch 7, In re A Debtor (No 68 of 1911) [1911] 2 KB

652, and In re Pitchford [1924] 2 Ch 260.

89. In my view, by becoming a party to legal proceedings in this jurisdiction, a person is brought within a system governed by rules of court, which carry with them the potential for being rendered legally liable for costs, subject of course to the discretion of the court. An order for costs made against a company in liquidation, made in proceedings begun before it went into liquidation, is therefore provable as a contingent liability under rule 13.12(1)(b), as the liability for those costs will have arisen by reason of the obligation which the company incurred when it became party to the proceedings.

90. I have little concern about overruling those earlier decisions, although they are long-standing. First, the judgments are very short of any reasoning, and consist of little but assertion. Secondly, they were decided at a time when the legislature and the courts were less anxious than currently for an insolvency to clear all the liabilities of a bankrupt (as they were all concerned with individual insolvencies). Although most of the provisions of rule 13.12 and section 382 can be found in section 30(3), (4) and (8) of the Bankruptcy Act 1914, over the past three hundred years, “the legislature has progressively widened the definition of provable debts and narrowed the class of non-provable liabilities” to quote from the written case of Mr Phillips QC who relied on those cases. Thirdly, those cases are impossible to reconcile logically with the earlier case of In re Smith, Ex p Edwards (1886) 3 Morrell 179, where, on identical facts (save that it was an arbitration rather than litigation) it was held that an order for costs did give rise to a provable debt. Fourthly, the unsatisfactory nature of those decisions can be seen from the way in which the Court of Appeal sought to evade their consequence in Day v Haine [2008] ICR 1102, a case which I consider to have been rightly decided.

91. For the same reasons, I consider that the decisions of the Court of Appeal in Glenister v Rowe [2000] Ch 76 and Steele [2006] 1 WLR 2380 were wrongly decided, although I can see how it might be said that they were justified on the basis of stare decisis. The reasoning of Arden LJ in the latter case at paras 21-23 is instructive, because, as she says, the previous authorities in relation to provable debts suggested a “narrower meaning of contingent liability” than was adopted by the majority in Sutherland. That observation neatly illustrates why they were wrongly decided.

92. The Report of the Review Committee on Insolvency Law and Practice (“the Cork Report”, 1982, Cmnd 8558), para 1289, described it as a “basic principle of the law of insolvency” that “every debt or liability capable of being expressed in money terms should be eligible for proof” so that “the insolvency administration

should deal comprehensively with, and in one way or another discharge, all such debts and liabilities”.

93. The notion that all possible liabilities within reason should be provable helps achieve equal justice to all creditors and potential creditors in any insolvency, and, in bankruptcy proceedings, helps ensure that the former bankrupt can in due course start afresh. Indeed, that seems to have been the approach of the courts in the 19^th century before the somewhat aberrant decisions referred to in para 88 above. Thus, in Ex p Llynvi Coal and Iron Co; In re Hide (1871) LR 7 Ch App 28, 32, James LJ described one of the main aims of the bankruptcy regime as to enable the bankrupt to be “a freed man – freed not only from debts, but from contracts, liabilities, engagements and contingencies of every kind”. If that was true in 1871, it is all the more true following the passing of the 1986 and 2002 Acts, and as illustrated by the amendment to rule 13.12(2) effected following the decision in In re T & N Ltd [2006] 1 WLR 1728, so as to extend the rights of potential tort claimants to prove.

94. It was suggested that para (m) was included in rule 4.218(3) on the assumption that cases such as those mentioned in para 88 above were rightly decided. That may be so. But, even if it is, the fact that a rule has been drafted on the basis that a decision of the Court of Appeal was right does not mean that this court should uphold the decision if satisfied that it was wrong.

Conclusion on the provable debt issue

95. I would accordingly dismiss these appeals to the extent of holding that the administrators are bound to meet the liabilities of the Target companies under the FSD regime, but allow the appeals to the extent of holding that these liabilities are to be treated as provable debts.

96. I have had the opportunity of reading in draft the judgment of Lord Sumption on this issue and I agree with it. His reference to Carron Iron Co Proprietors v Maclaren (1855) HL Cas 416 and In re Oriental Inland Steam Co (1873-4) LR 9 Ch App 557 appears to me to be very much in point. I have also found assistance in the careful judgment of Lord Drummond Young in the Outer House in In re Thomas v Burton, liquidator of Ben Line Steamers Ltd [2010] CSOH 174; 2011 SLT 535.

Is the liability under a FSD issued after an insolvency event a liquidation expense?

97. Given that the potential FSD liability in each of these cases is a debt falling within rule 13.12(1)(b), and therefore a provable debt within rule 12.3, and the acceptance on all sides that it would not therefore be an expense, it is strictly unnecessary to consider this question. However, it should be addressed as it was fully debated, and the point is of some potential importance, as I do not entirely agree with the courts below as to the effect of the reasoning and decision of the House of Lords in In re Toshoku Finance UK plc [2002] 1 WLR 671.

98. The question which would arise if the potential liability under the FSD regime did not give rise to a provable debt under rule 13.12 where the FSD is issued after the relevant insolvency event, is whether the liability would be within the expression “charges and other expenses incurred in the course of the … administration” within rule 12.2, and, more particularly, within the expression “any necessary disbursements by the administrator in the course of the administration”, within rule 2.67(1)(f) – the equivalent provision in a liquidation being rule 4.218(3)(m).

99. The word “necessary” in rule 2.67(1)(f) carries with it a legal obligation to pay (or, possibly, in exceptional cases, a moral obligation to pay, as to which see the next section of this judgment). However, that is somewhat circular, as it leaves open the very question which has to be decided, namely whether the liability in question which has been imposed on the company is one which the administrator must pay. Further, a liability may arise during an administration without falling within rule 2.67(1)(f), without being “in the course of” the administration. In Davidson v Robb [1918] AC 304, 321, Lord Dunedin explained that “in the course of his employment” had a more limited meaning than “during the period of his employment” and connoted “something which is part of his service” namely “work or the natural incidents connected with the class of work”, a view echoed by Lord Russell in Alderman v Great Western Railway Co [1937] AC 454, 459.

100. While it would be dangerous to treat any formulation as an absolute rule, it seems to me, at any rate subject to closer examination of the authorities and counter-arguments, a disbursement falls within rule 2.67(1)(f) if it arises out of something done in the administration (normally by the administrator or on the administrator’s behalf), or if it is imposed by a statute whose terms render it clear that the liability to make the disbursement falls on an administrator as part of the administration – either because of the nature of the liability or because of the terms of the statute.
101. Thus, if an administrator, on behalf of the company, enters into a transaction which gives rise to tax, or starts (or adopts) proceedings which give rise to a liability for costs, that tax or those costs would fall within the rule, as they arise from his actions as administrator during the administration. This conclusion is consistent with the authorities on liquidations – see eg In re Beni-Felkai Mining Co Ltd [1934] 1 Ch 406, In re Mesco Properties Ltd [1979] 1 WLR 558, affirmed [1980] 1 WLR 96 (tax), In re Trent & Humber Shipbuilding Co; Bailey & Leetham’s Case (1869) LR 8 Eq 94 and In re Wenborn & Co [1905] 1 Ch 413 (costs).

102. An area in which liquidators have been held liable to meet a statutorily imposed liability is that of property taxes – ie business and domestic rates and community charge – see, for instance, In re International Marine Hydropathic Co (1884) 28 Ch D 470, In re National Arms & Ammunition Co (1885) 28 Ch D 474, In re Blazer Fire Lighter Ltd [1895] 1 Ch 402, and more recently Exeter City Council v Bairstow [2007] Bus LR 813. The explanations in the judgments in those cases of the basis on which a liquidator has been held liable for rates and (in Kentish Homes) for community charge as an expense of the liquidation, are not entirely consistent. Sometimes it was said to be because the liquidator is retaining the property in question for the benefit of the winding-up – see eg per Baggallay LJ in Marine Hydropathic at 471 and Fry LJ in National Arms at 481. However, it was also said that the rates should count as an expense on the ground that, because the liquidator remained in rateable occupation of the property in question, the rates for the period should rank as an expense of the liquidation – see eg per Bowen LJ in National Arms at 480 and 482, and Vaughan Williams J in Blazer at 406-7.

103. The latter rationale seems to me to represent the current state of the law – see per Lord Hoffmann in Toshoku at para 34 and per David Richards J in Exeter at paras 15-19. In my view, therefore, the fact that the liability for rates falling due after an insolvency event on property retained by the liquidator ranks as an expense of the liquidation, is based on the proposition that, as a matter of interpretation, the rating (and community charge) legislation imposes such a liability on the liquidator (and the same logic must apply in an administration). This is consistent with the fact that liability for rates (and community charge), arises from day to day, and the liability is treated as an expense only in respect of the company’s occupation of property during the liquidation.

104. This conclusion derives a degree of support from the fact that, in the context of a liquidation, it is always open to a liquidator to disclaim onerous property (under sections 178-182 of the 1986 Act) and, if he chooses not to do so, it would presumably be as a result of a conscious decision to retain the property for the benefit of the creditors. An administrator cannot disclaim property, but there is force in the point that the rating authorities should not be worse off because a company opts for administration rather than liquidation, given that the normal

reason for preferring administration to liquidation is to seek a better outcome for creditors and/or shareholders of the company – see para 3(1) of Schedule B1 to the 1986 Act.

105. Adopting the approach I have suggested, it appears to me that a potential liability under a FSD or a liability under a CN does not fall within the scope of expenses of an administration within rule 12.2 or rule 2.67(1)(f). First, there is no question of such a liability resulting from any act or decision taken by or on behalf of the administrator or any act or decision taken during the administration. The liability self-evidently arises out of events which occurred before the insolvency event.

106. Secondly, I do not consider that the terms of the 2004 Act, properly interpreted, mean that a liability under a CN would be an expense of the administration, if it was not a provable debt under rule 13.12. It is true that the effect of a CN under section 49(3) of the 2004 Act is that it gives rise to a debt payable by the target once it is issued, but it does not seem to me that that can be sufficient to render the payment of the debt a “necessary disbursement … by the administrator in the course of the administration”. The mere fact that an event occurs during the administration of a company which a statute provides gives rise to a debt on the part of the company cannot, of itself, be enough to render payment of the debt an expense of the administration. It would be a debt payable “during the period of” the administration, but it would not be “part of” the administration, or a payment which was one of the “natural incidents connected with” the administration, to use the language of Lord Dunedin in Davidson.

107. In my view, something more would be required, either from the wording of the 2004 Act or from the nature of the liabilities which it imposes, before a CN issued after the target’s insolvency event could be held to be an expense of the administration or liquidation. The 2004 Act and the FSD Regulations are silent on the issue of the status of the liability under the FSD regime where the target has suffered or suffers an insolvency event. It is therefore necessary to consider whether there is any indication that can be gathered from the 2004 Act, its aims and procedures, that it was intended that such a liability should rank as an expense of the target’s administration or liquidation, if it does not give rise to a provable debt.

108. For the reasons given in paras 59-62 above, it would be remarkable if a liability under a CN issued to a target pursuant to a FSD issued after the target suffered an insolvency event had priority over the target’s other unsecured creditors, when a CN, based on precisely the same facts, would not have such priority if it was issued pursuant to a FSD issued notice issued before the insolvency event. I accept that it would be curious if a FSD issued after an

insolvency event was significantly less effective than one issued before that event, and indeed that it would be unlikely to result in any sum being paid as explained in para 63 above. However, the notion that liability pursuant to a FSD issued after the target had suffered an insolvency event ranks behind the target’s provable debts, is, to my mind, less surprising than the notion that it should rank ahead of them.

109. First, I consider that the balance of anomalies, as discussed in paras 59-63 above, is such that the former appears a much less unlikely outcome. Secondly, as a general proposition, once the facts giving rise to a right to raise a claim (in these cases, by issuing a FSD) exist, it would be very unusual for the beneficiary of the right to be better off as a result of a delay in raising the claim, but it would be far from surprising if the beneficiary were worse off as a result of such a delay.

110. The reason that the courts below reached a different conclusion is best explained by quoting a passage, from the first instance judgment, which the Court of Appeal (in paras 99-101 of Lloyd LJ’s judgment) expressly approved. At [2011] Bus LR 766, para 146, Briggs J said that Lord Hoffmann’s speech in Toshoku established as “a general rule” that:

“[W]here by statute Parliament imposes a financial liability which is not a provable debt on a company in an insolvency process then, unless it constitutes an expense under any other sub-paragraph in the twin expenses regimes for liquidation and administration, it will constitute a necessary disbursement of the liquidator or administrator. That is the general rule, whether the statute expressly refers to companies in an insolvency process as being subject to the liability, or whether the statute achieves the same result by using a criterion for liability which is insolvency neutral. Any other conclusion would in my judgment attribute an excessive weight to the linguistic method by which different legislation achieved the same result, namely that the statutory obligation in question is a liability of a company in an insolvency process.”

111. While it is fair to say that some observations of Lord Hoffmann in Toshoku, if read on their own, may appear to support that “general rule”, I consider that Briggs J’s summary amounts to an incorrect statement of the law. In my view, the general guidance given by Lord Hoffmann in Toshoku is to be found in para 46, where he said that “the question of whether [any particular] liabilities should be imposed upon companies in liquidation is a legislative decision which will depend upon the particular liability in question”. In a case, such as the present, where (i) the statutory liability is one which could have been imposed before or after liquidation, (ii) the liability does not give rise to a provable debt (as is being assumed for present purposes) and (iii) the statute is completely silent as to how

the liability should be treated if it is imposed after an insolvency event, the liability can only be an expense of the liquidation or administration if the nature of the liability is such that it must reasonably have been intended by the legislature that it should rank ahead of provable debts. It would be wrong to suggest that this is a test which may not need to be refined in future cases, but it appears to me to be supported by the facts and arguments raised on these appeals.

112. I do not consider that Toshoku takes matters any further in the present case. Lord Hoffmann explained in para 2 that the liability in Toshoku arose from a statutory provision which stated that:

“a company is ‘chargeable to corporation tax on profits arising in the winding up of the company’. It may be assessed in respect of an accounting period deemed to commence on the liquidation date …, and the liquidator is the proper officer liable to pay the tax …. [Other relevant statutory requirements were that] profits … must be computed on an accruals basis … [and] the computation must be made on the assumption that ‘every amount payable under the relationship will be paid in full as it becomes due’”.

In other words, unlike the present cases, Toshoku concerned a tax liability which was imposed on a liquidator, as opposed to the company, and it was a tax which only applied (in specified circumstances) to a company which had gone into liquidation.

113. As Lord Hoffmann said at para 30, “[t]here would be little point in a statute which specifically imposed liabilities upon a company in liquidation if they were payable only in the rare case in which it emerged with all other creditors having been paid.” Even in such a case, I consider that it would be appropriate for a court to consider whether the legislature intended the liabilities concerned to rank as an expense, but the point made by Lord Hoffmann would clearly be a very powerful factor as to why it should.

114. I therefore would conclude that, if the liability in these cases did not rank as a provable debt, it would not count as an expense of the administration.

Does the court have a residual discretion?

115. If I had taken a different view on the provable debt issue, an alternative argument to that just discussed was that the court has the power to direct the administrator of a Target company to accord to the potential liability under the

FSD regime a higher ranking than it would be given under the 1986 Act and the Insolvency Rules. In other words, that the court could order the administrator to treat the potential FSD liability as a provable debt (category 5 in para 39 above) even though the effect of the legislation is that it should rank lower (namely category 7).

116. At any rate at first sight, it would be extraordinary if a court, which had decided that a liability did not fall within the definition of provable debts in rule 13.12, could nonetheless go on to decide that it was to be so treated, in the absence of any specific statutory power to do so. Such a course would appear to be wrong in principle, because it would involve a judge effectively overruling the lawful provisions of a statute or statutory instrument. It would also be highly problematic in practice because it would throw many liquidations and administrations into confusion: the law would be uncertain, and many creditors who felt that the statutory ranking caused them unfair prejudice would make applications to the court.

117. If further reasons were required for this conclusion, they may be found in rule 2.67 and in Toshoku. Rule 2.67(2) and (3), referred to in para 42 above, show that, where the Insolvency Rules wish to give the court the ability to change the priority rules, they say so. In the course of his speech in Toshoku at para 38, Lord Hoffmann referred to the proposition “whether debts should count as expenses of the liquidation is a matter for the discretion of the court” and held that there was no such discretion and disapproved Sir Donald Nicholls V-C’s comments in In re Kentish Homes Ltd [1993] BCLC 1375. As Lord Hoffmann made clear in para 41, how a particular liability was to be ranked depended solely on the proper interpretation of the Insolvency Rules.

118. The justification for a contrary view was based on three paragraphs of Schedule B1 to the 1986 Act (“Schedule B1”), and a number of decisions where the court has ordered a liquidator to take a particular action.

119. Para 13 of Schedule 1 entitles an administrator to make any payment which is “necessary or incidental” to the performance of his functions. I do not see how that can entitle him, let alone the court to direct him, to treat an unprovable debt as a provable debt (unless, conceivably, there was resulting benefit which would redound for the benefit of the proving creditors, although even then it would be problematic). It can scarcely be said to be “incidental” or “necessary” to a person’s statutorily prescribed functions to do something inconsistent with those functions.

120. Para 65(3) of Schedule B1 precludes an administrator from paying a creditor who is neither secured nor preferential without the sanction of the court. I

cannot see how this provision can be properly interpreted as giving the court a roving commission to change the statutory priorities in a particular case simply because it does not like the consequences of those priorities. It was no doubt intended to apply where the payment in question is necessary or desirable to achieve one of the administrator’s statutory functions under paragraph 3 of Schedule B1 to the 1986 Act (eg the company’s survival or a more advantageous realisation of the company’s assets).

121. Para 74 of Schedule B1 entitles a creditor to apply to the court if it considers that the administrator proposes to act in a way which would unfairly prejudice it. This cannot, in my view, apply to a case where the administrator is proposing to do that which the legislation requires him to do. It applies where the administrator is exercising a power, or discretion, most obviously carrying on the company’s business in a certain way or selling off an asset of the company, or not performing an obligation, such as paying off creditors in the order mandated by the legislation. Again, it cannot have sensibly been intended to give the court a roving commission to vary the clear statutory ranking of liabilities as summarised in para 39 above.

122. As to the common law, there are a number of cases, starting with In re Condon Ex p James (1874) LR 9 Ch App 609, in which a principle has been developed and applied to the effect that “where it would be unfair” for a trustee in bankruptcy “to take full advantage of his legal rights as such, the court will order him not to do so”, to quote Walton J in In re Clark (a bankrupt) [1975] 1 WLR 559, 563. The same point was made by Slade LJ in In re TH Knitwear (Wholesale) Ltd [1988] Ch 275, 287, quoting Slater J in In re Wigzall, Ex p Hart [1921] 2 KB 835, at 845: “where a bankrupt’s estate is being administered … under the supervision of a court, that court has a discretionary jurisdiction to disregard legal right”, which “should be exercised wherever the enforcement of legal right would

… be contrary to natural justice”. The principle obviously applies to administrators and liquidators – see In re Lune Metal Products Ltd [2007] 2 Bus LR 589, para 34.

123. However, none of these cases begins to justify the contention that an administrator can be ordered to change the ranking of a particular debt simply because the statutory ranking appears unattractive – in this case because it means that a particular debt is ranked lower than other unsecured debts because (as I am assuming) it is not provable according to the statutory formula. Indeed, observations in Lune Metal, paras 35-38, tend to support the notion that the court cannot sanction a course which would be outside an administrator’s statutory powers.

124. It is right to mention that the court has sanctioned an otherwise unauthorised payment where a company in administration wishes to avoid the cost

of going into compulsory liquidation. In such cases, which include In re UCT (UK) Ltd [2001] 1 WLR 436 and Lune Metal, the terms on which the administrators are discharged includes a direction which ensures that the preferential creditors are in no worse a position than if there had been a compulsory liquidation. However, those cases provide no assistance to the argument that the court can direct a FSD regime liability to be “promoted” ahead of its statutory ranking, as (i) the direction benefits the creditors with provable debts, (ii) the direction is tied to the discharge of the administrators, not the performance of their on-going functions, and (iii) the direction does not involve any conflict with statutory ranking of claims, and in particular, it does not harm the interests of the creditors’ with provable debts: on the contrary, they benefit from the direction.

125. In the present cases, I understand the attraction of the argument that the court should order the administrators of the Target companies to treat the potential FSD liabilities as provable debts (if they are not so provable): otherwise, they may be valueless. However, I come back to the point that, if the effect of the Insolvency Rules is that the liabilities are not provable debts, there is no basis for the court deciding that they are. It would be wrong for the courts to override the statutory ranking, especially given it would cause significant prejudice to others (in this case the creditors with provable debts). That is particularly true as the liabilities are statutory, so that the legislature could have dealt with their status in a liquidation or administration (as indeed it did in relation to a section 75 debt).

126. The argument to the contrary also relied on the fact that the liability of a target under the FSD regime would have been a provable debt if the FSD had been issued before the insolvency event. That is undoubtedly an argument in favour of the liability where the FSD is issued after an insolvency event being a provable debt, as mentioned above. However, if the liability is not a provable debt in such circumstances, the argument does not support the contention that the administrator can be required to treat it as if it were. As already mentioned, the mere fact that the court does not think it fair that a particular statutory liability should not rank as a provable liability under the relevant statutory provisions is not enough to justify a decision to alter the effect of those provisions.

127. The point can be taken a little further. The decision of the courts below, that the liability in these cases was an expense of the administration and not a provable debt, was unattractive for the reasons given in paras 59-62 above. It seems to me that, if, as is suggested by the argument I am considering, the courts had had power to do so, they should have gone on to hold that it would nonetheless direct the administrators to treat the liability as a provable debt. Such a direction would not merely have been a surprising one, but it would have been one which flew in the face of Lord Hoffmann’s observations at paras 38-41 in Toshoku, disapproving In re Kentish Homes Ltd [1993] BCLC 1375.

Conclusion

128. I would accordingly allow these appeals to the extent of declaring that a Target company’s liability under the FSD regime, arising pursuant to a FSD issued after the company has gone into administration, ranks as a provable debt of the company, and does not rank as an expense of the administration.

LORD SUMPTION (with whom Lord Mance and Lord Clarke agree)

129. I agree with the order proposed by Lord Neuberger and with his reasons. I add a few observations of my own on a point which might be regarded as a matter of wholly abstract jurisprudence if it were not fundamental to the analysis of the effect of this particular scheme.

130. The critical question is what constitutes an “obligation incurred” for the purpose of rule 13.12(1)(b) of the Insolvency Rules 1986. The context shows it means a legal rule applying before the date when the company goes into liquidation which may, contingently on some future event, give rise to a “debt or liability” arising after that date. But it cannot extend to every legal rule which may on any contingency have that effect. Otherwise every debt or liability would be provable irrespective of the date when it accrued, unless the law changed after the company went into liquidation. Since the scheme depends on there being a common date as at which the fund falls to be valued and distributed pari passu, that cannot be right. Some limitation must be read into sub-paragraph (b). But what limitation?

131. The paradigm case of an “obligation” within the sub-paragraph is a contract which was already in existence before the company went into liquidation. It is implicit in the argument of those who contend on this appeal that there is no provable debt, in this case that contract is not just the paradigm case but the only one. Yet when one asks what it is about a contract that qualifies it as a relevant source of obligation, the answer must be that where a subsisting contract gives rise to a contingent debt or liability, a legal relationship between the company and the creditor exists from the moment that the contract is made and before the contingency occurs. The judgment of Lord Reid in In re Sutherland (dec’d) [1963] AC 235 was concerned with a very different statutory scheme, but his analysis is nevertheless illuminating because it makes precisely this point at pp 247-8:

“It is said that where there is a contract there is an existing obligation even if you must await events to see if anything ever becomes payable, but that there is no comparable obligation in a case like the

present. But there appears to me to be a close similarity. To take the first stage, if I see a watch in a shop window and think of buying it, I am not under a contingent liability to pay the price: similarly, if an Act says I must pay tax if I trade and make a profit I am not before I begin trading under a contingent liability to pay tax in the event of my starting trading. In neither case have I committed myself to anything. But if I agree by contract to accept allowances on the footing that I will pay a sum if I later sell something above a certain price I have committed myself and I come under a contingent liability to pay in that event.”

132. Contract is not the only legal basis on which a contingent obligation of this kind may arise. A statute may also give rise to one. A good example is the substantive obligation which English law has always held to be owed by a debtor under a foreign judgment. It is the basis of the common law action to enforce it. Another is the obligation of a creditor arising from the statutory scheme of distribution in an English insolvency, not to seek by litigation in a foreign court a priority inconsistent with that scheme: see Carron Iron Co Proprietors v Maclaren (1855) HL Cas 416, 440 per Lord Cranworth LC, In re Oriental Inland Team Co (1873-4) LR 9 Ch App 557, and in the United States Cole v Cunningham (1882) 133 US 107. In both of these examples, a legal relationship is created between the debtor and other persons, albeit without contract. In the first, it is the legal relationship with the judgment creditor arising from the fact that the judgment debtor was subject to the jurisdiction of the foreign court, whether by virtue of residence or submission. In the second, it is the legal relationship of the creditor with the debtor company and with other creditors arising from the statutory scheme of distribution. If the mandatory provisions of a statute may create a legal relationship between the company and a creditor (or potential creditor) giving rise to a provable debt, then there is no reason why it should not do so contingently upon some future event.

133. In In re Sutherland decd [1963] AC 235 the company’s liability for balancing charges by way of recoupment of capital allowances, which the majority held should be taken into account when valuing its assets, did not exist at the valuation date because at that date it was still contingent upon a future sale of those assets. It was nevertheless a relevant contingent liability for valuation purposes, because at the valuation date there was a legal relationship between the companies and the Crown arising from the statutory scheme which made capital allowances subject to balancing charges in the contingency of a sale.

134. In Secretary of State for Trade and Industry v Frid [2004] 2 AC 506 the Secretary of State was subrogated by statute to the claims of employees to compensatory notice pay and redundancy payments. The liability of the company to meet those claims did not arise until the employees were dismissed, which was

after the company went into liquidation. But the obligation existed before, because the statutory scheme superimposed upon the contract of employment created the legal relationship which made the compensatory notice pay and the redundancy payments due. Lord Hoffmann, with whom the rest of the committee agreed, said at paras 17, 19:

“if the Secretary of State had agreed by contract before the insolvency date to guarantee any future liability of the company to pay compensatory notice pay or make redundancy payments to employees under the 1996 Act, the contract of guarantee would have created a contingent liability on the part of the company to reimburse the Secretary of State which was a ‘debt’ at the insolvency date and became capable of set-off when the employees were afterwards paid. The next question is whether it makes a difference that the contingent liability existed by virtue of a statute rather than a contract and, not being consensual, that it involved no direct contract or other relationship with the employees or the company… If a statutory origin does not prevent set-off in the case of debts due and payable at the insolvency date, I do not see why it should make any difference that the statute creates a contingent liability which exists before the insolvency date but falls due for payment and is paid afterwards.”

135. In the Victoria case of Lofthouse v Commissioner of Taxation [2001] 164 FLR 106, the statute conferred upon the Commissioner an indemnity against the directors of a company if tax payments under the Australian equivalent of PAYE were subsequently held repayable as insolvent transactions (in effect, preferences). The indemnity was contingent upon the tax being determined to be repayable after the employer had gone into liquidation, but the statutory scheme created the relevant legal relationship between the directors and the Commissioner as soon as the tax payments were made by the company. They were therefore provable as contingent debts in the insolvency of the directors. Warren J observed at p 118:

“The potential liability of the third parties in this proceeding is a contingent liability within the meaning of s 82(1) of the Act because the potential liability arose from an obligation pursuant to an indemnity. Furthermore, all the objective circumstances giving rise to the potential for the invocation of the chose in action represented by the right to indemnity had transpired prior to the third parties entering into their composition under Pt X of the Bankruptcy Act.”

136. In the present case, the Court of Appeal considered itself to be bound by a line of cases in which it was held that a liability for costs arising from a judgment given after the commencement of the insolvency was not provable as a contingent debt, even if the litigation was in progress when the company went into liquidation. The case-law begins with In re Bluck Ex p Bluck (1887) 57 LT 419, and continues with In re British Gold Fields of West Africa [1899] 2 Ch 7, In re A Debtor (No 68 of 1911) [1911] 2 KB 652, In re Pitchford [1924] 2 Ch 260, Glenister v Rowe [2000] Ch 76. The reasoning of these cases has recently been applied to other claims said to represent contingent liabilities: see R (Steele) v Birmingham City Council [2006] 1 WLR 2380. There are a number of problems about these cases. One of them, as it seems to me, is the absence of any real attempt to analyse the effect of the statutory scheme in creating an obligation to meet a liability contingently on some specified event. In the earlier cases, this can perhaps be regarded as the legacy of the older principle which admitted only contractual debts to proof. But that consideration cannot explain the more recent decisions. In my view they were wrongly decided. In the costs cases, I consider that those who engage in litigation whether as claimant or defendant, submit themselves to a statutory scheme which gives rise to a relationship between them governed by rules of court. They are liable under those rules to be made to pay costs contingently on the outcome and on the exercise of the court’s discretion. An order for costs made in proceedings which were begun before the judgment debtor went into liquidation is in my view provable as a contingent liability, as indeed it has been held to be in the case of arbitration proceedings: In re Smith, Ex p Edwards (1886) 3 Morrell 179. In both cases, the order for costs is made against some one who is subject to a scheme of rules under which that is a contingent outcome. The fact that in one case the submission is contractual while in the other it is not, cannot make any difference under the modern scheme of insolvency law under which all liabilities arising from the state of affairs which obtains at the time when the company went into liquidation are in principle provable. Of course, an order for costs like many other contingencies to which a debt or liability may arise, depends on the exercise of a discretion and may never be made. But that does not make it special. It is not a condition of the right to prove for a debt or liability which is contingent at the date when the company went into liquidation that the contingency should be bound to occur or that its occurrence should be determined by absolute rather than discretionary factors.