Harbinger Capital Partners v Caldwell (As the Independent Valuer of Northern Rock Plc) & Anor (Rev 1)

The troubles that beset Northern Rock in the autumn of 2007 were the harbinger of the near collapse of the global banking system in the following year. On 22 February 2008, following attempts to keep it going in the private sector, Northern Rock was nationalised. Upon nationalisation all existing shares in Northern Rock passed into Government hands. The issue on this appeal is the basis on which compensation for those shares is to be determined.

The compulsory transfer of shares in Northern Rock to the Treasury Solicitor, as nominee of the Treasury, was effected by article 2 of The Northern Rock plc Transfer Order 2008. This was an order made under powers contained principally in section 3 of the Banking (Special Provisions) Act 2008 (“the Act”). Section 5 of the Act required the Treasury, in relation to any order made under section 3 transferring securities to the public sector, to make a scheme for determining the amount of any compensation payable to the persons who held the securities immediately before the transfer. Section 5 (4) provided:

“In determining the amount of any compensation payable by the Treasury by virtue of any provision in an order under this section, it must be assumed—

(a)

that all financial assistance provided by the Bank of England or the Treasury to the deposit-taker in question has been withdrawn (whether by the making of a demand for repayment or otherwise), and

(b)

that no financial assistance would in future be provided by the Bank of England or the Treasury to the deposit-taker in question (apart from ordinary market assistance offered by the Bank of England subject to its usual terms).”

Financial assistance was in turn defined by section 15 (1) as follows:

““financial assistance”, in relation to any person, includes—

(a)

assistance provided by way of loan, guarantee or indemnity,

(b)

assistance provided by way of any transaction which equates, in substance, to a transaction for lending money at interest (such as a transaction involving the sale and repurchase of securities or other assets), and

(c)

assistance falling within paragraph (a) or (b) provided indirectly to or otherwise for the benefit of the person (including the provision of assistance within paragraph (a) or (b) to any group undertaking of that person),

whether provided in pursuance of an agreement or otherwise and whether provided before or after the passing of this Act.”

Pursuant to its statutory duty, the Treasury made The Northern Rock plc Compensation Scheme Order 2008. The scheme was set out in a schedule to the order. Paragraph 3 (2) of the schedule provided:

“The amount of compensation payable to a person shall be an amount equal to the value immediately before the transfer time of all shares in Northern Rock held immediately before the transfer time by that person.”

The transfer time was defined as the beginning of 22 February 2008. Paragraph 6 of the schedule said:

“In determining the amount of any compensation payable by the Treasury to any person in accordance with paragraphs 3 to 5, it must be assumed (in addition to the assumptions required to be made by section 5 (4) of the Act (compensation etc for securities transferred etc)) that Northern Rock

(a)

is unable to continue as a going concern; and

(b)

is in administration.”

The amount of the compensation was to be assessed by an independent valuer appointed by the Treasury. The appointee was Mr Caldwell. He determined the value of all shares in Northern Rock (both preference shares and ordinary shares) as nil. In so doing he applied the valuation assumption in section 5 (4) (a) on the basis that it required him to assume that at the transfer time Northern Rock had repaid all monies that had been lent to it by the Bank of England; and had realised sufficient of its assets to enable that to have been done. Harbinger Capital Partners, who had an interest in preference shares in Northern Rock, challenged that decision before the Upper Tribunal (Tax and Chancery Chamber). The challenge failed. The Tribunal (Mr Justice Warren P, Mr Andrew Bartlett QC and Ms Sandi O’Neill) held that the valuer was right in his interpretation of section 5 (4) (a). Their decision is at [2011] UKUT 408 (TCC), and is available on bailii. For the reasons that follow, in my judgment (although a minority one) the Tribunal was wrong.

The factual background is set out in the judgment of Laws LJ in SRM Global Master Fund LP v HM Treasury [2009] EWCA Civ 788. I pick out only a few salient points that are relevant to the current appeal. At all material times, both before and after the provision of financial assistance by the Bank of England and HM Treasury, Northern Rock’s assets as recorded in its balance sheet exceeded its liabilities. Its problem was liquidity: cash flow insolvency rather than balance sheet insolvency.

The Bank of England intervened as lender of last resort. The then Governor of the Bank of England explained the principles applicable in a lecture quoted by Laws LJ. The following are the relevant points for the purposes of this appeal:

“First, we will explore every option for a commercial solution before committing our own funds. Initially, we will always look to major shareholders to provide support. Short of that, we will encourage the bank to try to find a buyer...

Second, central banks are not in the business of providing public subsidy to private shareholders. If we do provide support, we will try to structure it so that any losses fall first on the shareholders and any benefits come first to us. And any support we provide will be on terms that are as penal as we can make them, without precipitating the collapse we are trying to avoid.

Third, we aim to provide liquidity: we will not, in normal circumstances, support a bank that we know at the time to be insolvent. Our own capital is not there to be used as risk capital. But it would be wrong to conclude from this that loans or guarantees never involve any risk...

Fourth, we look for a clear exit. The company may be required to run down or restructure its operations, under our surveillance, to the point where it can do without our support within a given period. Making the terms of our support as unattractive as possible has the great advantage of encouraging this process... We aim to protect the system, not to keep in being unviable banking capacity and so interfere in the market process unnecessarily.”

Assistance from the Bank of England and the Treasury was given in stages and took different forms. The assistance was not given for the purpose of rescuing Northern Rock as an end in itself, but for the protection of the banking system as a whole. Liquidity support was first given on 14 September 2007. The Treasury press announcement contained the statement that “The FSA judges that Northern Rock is solvent, exceeds its regulatory capital requirement and has a good quality loan book”. This is consistent with the third principle in the Governor’s lecture. Three days later the Chancellor of the Exchequer announced that the government would guarantee existing deposits in Northern Rock. The guarantee was extended in the following month to cover new retail deposits. The guarantee was paid for by a fee charged to Northern Rock. The guarantee was subsequently extended again to cover certain unsubordinated wholesale obligations. Ultimately it covered all unsecured retail products, all uncollateralised and unsubordinated wholesale products and wholesale borrowings, all payment obligations under uncollateralized derivative transactions and (to a limited extent) all collateralised derivatives and collateralised wholesale borrowings (including covered bonds). By 31 December 2007 the Bank of England had lent almost £27 billion to Northern Rock, and the Treasury had assumed contingent liabilities under guarantees to the tune of about £29 billion. These loans were secured in various ways.

Since the assistance was coming from the state, EU rules about state aid came into play. The European Commission has issued Community Guidelines on State Aid for Rescuing and Restructuring Firms in Difficulty (2004/C 244/02). They point out that rescue aid is by nature temporary and reversible assistance. Rescue aid must be limited to the minimum necessary for a period not exceeding six months. The aid must consist of reversible liquidity support in the form of loan guarantees and loans at a proper interest rate (§ 15). The guidelines go on to set out the criteria that must be satisfied in order for the Commission to approve the rescue aid (§ 25). These criteria include:

The government sought retrospective approval for the arrangements from the European Commission. The Commission’s decision authorising state aid to Northern Rock contained the following:

“The Commission expects your authorities to respect their commitment to communicate to the Commission, not later than 17 March 2008, a credible and substantiated restructuring plan or a liquidation plan or proof that the aid measures have been repaid in full and that the guarantees have been terminated.”

As the Tribunal pointed out (§ 153) this meant that:

“… it was incumbent on the authorities to receive actual repayment of any loans given by way of financial assistance on or after 17 September 2007 in the absence of either an appropriate restructuring plan or a liquidation plan.”

Our attention was also drawn to the Commission’s notice on the recovery of unlawful and incompatible State aid (2007/C 272/05). However, I do not consider that this bears on the case for three reasons. First, it is concerned with unlawful aid, and the aid given to Northern Rock was not unlawful. Second, the notice recognises that the liquidation of a recipient of State aid and the sale of its assets “under market conditions” is an acceptable means of recovery (§ 61). Third, the Commission’s position in relation to the crisis in the banking world was that, subject to regular review, repayment of authorised State aid could be spread over a period of two years compatibly with the rescue and restructuring guidelines (Communication from the Commission 2008/C 270/02).

Attempts were made to find a private sector buyer for Northern Rock, but the attempts failed to produce a satisfactory solution. The regulatory authorities had already canvassed the possibility that in that event Northern Rock might have to be taken into public ownership. A joint press statement of 21 January 2008 said:

“The principles for assessing compensation, which would be set out in the legislation brought forward, would reflect the principle that the Government should not be required to compensate shareholders for value which is dependent on taxpayers' support and the fact that public sector ownership would be an alternative to an administration of the company. Accordingly, the compensation would be assessed by the valuer on the basis, among other things that all financial assistance to Northern Rock from the Bank of England or HM Treasury (including HM Treasury's existing guarantee arrangements) had been withdrawn and no other financial assistance (apart from Bank of England assistance on its usual terms through standing facilities or open market operations) were made available by them to Northern Rock.”

As things turned out there was no private sector buyer for Northern Rock unless government support continued to be made available. On 17 February 2008 the Chancellor of the Exchequer announced that Northern Rock would be nationalised. In making his announcement the Chancellor said:

“The Financial Services Authority continue to assure me the bank is solvent. It believes that Northern Rock's mortgage book is of good quality.”

The Bill that became the Banking (Special Provisions) Act 2008 was introduced into Parliament on 19 February 2008. As is now the practice, it was accompanied by Explanatory Notes. The note to section 5 (4) said:

“The assumptions are that all financial assistance provided by the Bank of England or the Treasury has been withdrawn; and that no further public assistance would be provided to the deposit-taker (apart from ordinary market assistance on its usual terms). Any announcement by the Treasury that they would, if necessary, put guarantee arrangements in place would also be disregarded. These assumptions ensure that any value that is dependent on public support provided to the deposit-taker is disregarded when compensation is determined.”

The Act was passed two days later on 21 February 2008.

Once again the question of state aid arose. In their decision of 2 April 2008 the European Commission considered the valuation criteria that would apply to the assessment of compensation for shareholders, and concluded:

“If the shareholders are only compensated on the basis of an independent valuation of the company without any State support, the purchase of the shares from the existing shareholders does not constitute State aid.”

The issue in the appeal turns on the meaning of the assumption in section 5 (4) (a) of the Act that:

“all financial assistance provided by the Bank of England or the Treasury to the deposit-taker in question has been withdrawn (whether by the making of a demand for repayment or otherwise)”

As noted, the Tribunal upheld the valuer’s interpretation of that assumption. Harbinger say that that interpretation is wrong, and put forward two alternatives:

The valuation criteria in the Act and the scheme have not been spelled out in detail, and quite a lot of the detail needs to be filled in. The ultimate objective of the exercise is to ascertain the “amount equal to the value” of the shares at the transfer time. The amount, obviously, is a monetary amount measured in pounds sterling, even though the scheme does not say so. It is also inherent in the concept of ascribing a monetary value to an asset such as a share that what you are asking is how much someone would expect to receive for it at the relevant time. And that in turn necessitates postulating a transaction of sale and purchase on the valuation date and asking what the purchase price would have been (see Waters v Welsh Development Authority [2004] UKHL 19 [2004] 1 WLR 1304 § 15). There was, of course, no sale in fact, so that the postulated transaction is a hypothetical one.

There are many areas of the law in which an amount is to be ascertained by postulating a hypothetical transaction of one kind or another. Rating is perhaps the oldest example, for which purpose rateable value was measured by postulating the hypothetical grant of a tenancy from year to year. But hypothetical transactions abound in other areas of the law: for example compulsory acquisition, taxation and rent review clauses. Sometimes the hypothesis is statutory and sometimes it is contractual. The courts have developed a well-established set of principles that apply to both kinds of case. The most important of these is that things are to be taken as they are in reality on the valuation date, except to the extent that the instrument postulating the hypothetical transaction requires a departure from reality. In the old cases this is summarised in the Latin phrase rebus sic stantibus. In the more modern cases it has been described as the principle of reality: Hoare v National Trust (1998) 77 P & CR 366.

The following points amplify the reality principle:

On behalf of the valuer Mr Howard QC submitted that the reality principle had no part to play in the interpretation of the hypothesis. The hypothesis should be interpreted by reference to the legislative purpose. The reality principle only had relevance when the hypothesis (correctly interpreted) came to be applied. Mr Howard cited no authority in support of this proposition. I accept, of course, that the hypothesis must be interpreted in the light of the legislative purpose; but I do not accept that the reality principle has no part to play in interpretation. That would be inconsistent with the observations collected in [23] (i) above. In addition the reality principle was, for example, heavily relied on by Hoffmann LJ in interpreting the rent review clauses considered in Co-Operative Wholesale Society Ltd v National Westminster Bank plc [1995] 1 EGLR 97, and by the Court of Appeal in StMartins Property Investments Ltd v CIB Properties Ltd [1999] L & TR 1 in which Aldous LJ said (p 10):

“It [i.e. the rent review clause in that case] should be construed with reality in mind and unless the words are clearly to the contrary to reflect that reality.”

In the same case Buxton LJ began his consideration by saying (p 17):

“… it is important again to remind ourselves of what is meant in the language of this part of the law by the presumption of reality and the effect that that has upon the construction of rent review clauses. ”

In my judgment the principles I have summarised are those which we should apply in our case. Accordingly the value of the shares should be assessed by reference to the facts as they stood at the transfer time, except to the extent that either the Act or the scheme requires a departure from reality. A departure from reality must either be expressly required or must be an inevitable consequence of what has been expressly required. Where a departure from reality is required, that departure should be kept to a minimum. It is unfortunate that none of these principles (or the cases which establish them) was drawn to the attention of the Tribunal.

I turn, then, to reality at the transfer time.

Northern Rock’s balance sheet at the transfer time showed assets of £106.31 billion and liabilities of £104.69 billion resulting in shareholders’ funds of £1.63 billion. Of those liabilities £25.28 billion represented financial assistance that had been given by the Bank of England. The financial assistance took various forms.

First, there were Treasury guarantees. Second, there was a Bank of England repo agreement over Treasury investments. Third, there was a loan secured by a fixed charge over a pool of residential mortgages. Fourth, there was a loan secured by a fixed and floating charge over those of Northern Rock’s assets as had not already been encumbered. All the loans were payable on demand, as is the Bank of England’s normal practice when acting as lender of last resort.

As at the transfer time none of these loans had been called in; and, necessarily, none had been repaid. Nor had Northern Rock realised assets with a view to repayment of the loans.

It is also important to emphasise that the assistance that the Bank of England had provided was liquidity support; that is to say the Bank’s intervention enabled Northern Rock to pay its debts as they fell due. The Bank’s intervention was not designed to shore up Northern Rock’s balance sheet. That would have been incompatible with the principles on which the Bank acts as lender of last resort. There is no evidence or finding that Northern Rock’s balance sheet was increased as a result of public sector support. As far as the evidence goes Northern Rock had a surplus of assets over liabilities when the Treasury first intervened, when the Bill that became the Act was introduced into Parliament and at the transfer time itself. The consequence of liquidity support was that Northern Rock was enabled to continue as a going concern. Without that support, as was common ground, it would have had to have entered administration or liquidation. But it would have done so on the basis that the starting point of the administration or liquidation was a company whose assets, as shown in its balance sheet, exceeded its liabilities.

The valuer set out his interpretation of the valuation assumptions, and their effect on value in the Consultation Document that he prepared in December 2009. He said (§ 2.30):

“I propose to proceed on the basis that the withdrawal of the BoE funding is achieved by the realisation of assets immediately prior to the Valuation Date, i.e. outside the scope of the assumed administration. It is clear that in the context of the Valuation Assumptions there is likely to have been a marked discount to the book value of the majority of those assets on any realisation. This is because of the:

•

requirement for immediate realisation;

•

likelihood that there would be a limited number of buyers (with lack of available funding) and no ready market; and

•

perception of a distressed sale.”

Although Northern Rock had some liquid assets (e.g. cash and gilts) the valuer concluded that the majority of the assets that would have to be sold were residential mortgages. Having considered evidence of sales of loan books he concluded that Northern Rock would have had to have accepted (on average) a 15 per cent discount in order to effect a sale of sufficient mortgages to repay the Bank of England. In considering which assets Northern Rock were assumed to have sold, the valuer made yet another assumption:

“I propose to assume that the best quality assets are realised and that the remaining assets on the balance sheet are of lower quality as they have more inherent risks.”

The impact of this approach was to turn Northern Rock from being a company which was balance sheet solvent (which it had been in reality both when the Bank of England first intervened and also at the transfer time) into a company which, in the hypothetical world, was balance sheet insolvent. As a result of the assumptions that the valuer made the balance sheet value of shareholders’ funds was reduced from a surplus of about £1.6 billion to a deficiency of about £2.4 billion.

The valuer’s interpretation, which the Tribunal accepted, contains within it the following counter-factual assumptions:

All these assumptions are departures from reality. Which of them (if any) are required by the terms of the Act or the scheme? There are five interconnected strands in my thinking:

Although there was debate about the legislative purpose behind section 5 (4) (a), in the end I do not think that there was any disagreement. The purpose of section 5 (4) (a) was to ensure that value dependent on public support was eliminated. This purpose was also the purpose underpinning the EU legislation to which we were referred. But there is no finding by the Tribunal that Northern Rock was or remained balance sheet solvent as a result of public support. Indeed the evidence shows clearly, to my mind, that Northern Rock was balance sheet solvent both before and after the Bank of England’s intervention. Nor is there any finding by the Tribunal that Northern Rock’s assets were increased by reason of the Bank of England’s intervention. The public support enabled Northern Rock to continue as a going concern; but it is common ground that the inevitable consequence of the assumptions in section 5 (4) is that Northern Rock could no longer continue as a going concern in the hypothetical world but would, instead, have entered administration. Thus the legislative purpose both domestically and at the European level would appear to be satisfied without having to assume that Northern Rock had realised assets when in fact it had not.

Mr Howard submitted that section 5 (4) should be interpreted conformably with EU law, even if that meant adopting an interpretation that was not the most obvious or natural one. As a general proposition I agree: Lister v Forth Dry Dock and Engineering Co Ltd [1990] 1 AC 546. What I found more difficult to grasp was what breach of EU law would be occasioned by the interpretation advanced by Mr Phillips QC for Harbinger. In the real world the European Commission had approved the State aid that had been given to Northern Rock; and had also approved the terms of nationalisation. So there was in fact no breach of EU law. In the hypothetical world, it must be assumed that all State aid has been withdrawn and that no more will be forthcoming. Moreover, under EU law where State aid has been sanctioned by the Commission, a restructuring or liquidation plan may be an acceptable alternative to repayment. So in the hypothetical world there is no breach of EU law either. Mr Howard suggested that it would be a breach of EU law if compensation to shareholders were to be assessed on the basis that the Bank of England’s loans remained “embedded” in Northern Rock for the duration of the hypothetical administration. But as he himself pointed out in the real world the loans were repayable on demand with a back-stop date of March 2008; and the Bank of England had ample security for the purpose of enforcing repayment of the loans. There is no reason in law to assume that the loans remained “embedded” in Northern Rock. Indeed in my judgment there is no legal justification for making assumptions about the future at all. What the valuer has to do is to assess a potential purchaser’s perceptions of all future possibilities on the valuation date. In assessing future possibilities the purchaser will have regard to current facts. Among the facts current at the transfer time was the Bank of England’s policy for recovering loans. There is no more reason to ignore that real fact than there was for ignoring the National Trust’s real policy for acquiring historic houses in Hoare v National Trust. If the Bank of England’s policy in the real world was to seek immediate recovery of the loans and if, in the hypothetical world, Northern Rock’s loans had been called in, it is for the valuer to assess the impact of that on the mind of a hypothetical purchaser of Northern Rock’s shares.

Both parties concentrated their submissions to the Tribunal and their written arguments in this court on the meaning of “withdrawn” in section 5 (4) (a). I prefer to start a little earlier and consider the meaning of “financial assistance”. It seems to me that inherent in the idea of “assistance” is that the assistance is something that is voluntary on the part of the person giving the assistance. Assistance is something that is given; not just taken. This, to my mind, is reinforced by the other part of the phrase “assistance… provided”. If a thief steals my wallet he may be able to take the benefit of the (little) money that he finds in it, but it would be a misuse of language to say that I had provided him with financial assistance. Likewise if I buy goods on 30 days’ credit it is clear that the seller has given me 30 days’ financial assistance. But if I fail to pay after 30 days, and the seller immediately protests, it seems to me to be difficult to say that the seller has provided me with more than 30 days’ financial assistance. My first impression, therefore, is that what is to be assumed to have been withdrawn is the assent of the person who had provided the assistance to its continuation.

I consider that this first impression is reinforced by the words in parenthesis: “(whether by the making of a demand for repayment or otherwise)”. The parenthesis is, in my judgment, intended to illustrate how financial assistance may be withdrawn. One way is by making a demand for repayment. This seems to me to be the clear meaning of the parenthesis. Financial assistance can be withdrawn “by” making a demand. In other words, the making of a demand is the means of withdrawing financial support. Mr Howard submitted (and the Tribunal accepted) that the words in parenthesis were put in out of an abundance of caution and served to emphasise that the precise mechanics of the withdrawal were immaterial. I am inclined to agree that the precise mechanics of withdrawal are immaterial, but that does not confront the question: what were the mechanics of withdrawal? I see no escape from the conclusion that, as a matter of the ordinary use of English, at least one method of effecting a withdrawal is “by making a demand for repayment”. The valuer’s interpretation either involves reading the parenthesis as if it said “whether initiated by making a demand for repayment or otherwise”; or, alternatively as reading it as if it said “whether by repayment or otherwise”, thus ignoring the reference to the demand entirely.

It follows, in my judgment, that the Act itself provides clear authority for the counter-factual assumption that the Bank of England has made a demand for repayment of its loans. If I am right in thinking that financial assistance is withdrawn “by” making a demand for payment, then it follows that the assumption is satisfied by assuming that such a demand has been made. However, I do not find within those words clear authority for any further counter-factual assumption. The Tribunal considered whether a loan could be said to have been withdrawn unless it had been repaid. But in my judgment that is not the right question. The statutory hypothesis is not that the loan has been withdrawn. The statutory hypothesis is that the assistance provided by way of loan has been withdrawn. The loan was the means by which (historically) the assistance was provided. The phrase is merely descriptive of the kind of assistance that must be assumed to have been withdrawn. Moreover, an assumption that the loans have in fact been repaid is, to my mind, at odds with the assumption that a demand for repayment has been made. A demand “for repayment” looks forward to a future repayment of the debt demanded; but it does not necessarily entail an immediate repayment. Moreover, as Beatson LJ pointed out in argument, the EU background papers speak consistently of the “recovery” of State aid. Had the draftsman wanted a portmanteau word that would cover all forms of State aid (i.e. not restricted to the deemed repayment of loans) there was one readily to hand. The Act could thus have said that it was to be assumed that all aid had been recovered, but it did not. Mr Howard was unable to give any example of financial assistance that, on the valuer’s interpretation of section 5 (4) (a), could actually be withdrawn “by making a demand for payment”. As he put it: “there would always be something more to do”. It must follow that the valuer’s interpretation gives no effect at all to these words in section 5 (4) (a). That is a very unusual way to interpret an Act of Parliament.

That a debtor repays a debt immediately upon a demand for repayment is a possible consequence of the demand; but I do not think that it is a necessary one. It is equally possible that a debtor, particularly one who is in financial trouble, delays payment for as long as possible.

Mr Howard also made brief submissions about the effect of the assumption on a repo agreement. Repo agreements are included in the definition of financial assistance in section 15 (1) of the 2008 Act (“assistance provided by way of any transaction which equates, in substance, to a transaction for lending money at interest (such as a transaction involving the sale and repurchase of securities or other assets)”). His point was that in the case of a repo agreement the Bank of England could simply sell the securities to which it had title; and that this was inconsistent with Harbinger’s position that the loans had been demanded, but not repaid. As I understand a repo agreement is an agreement for the sale of securities together with an agreement for the seller to buy back the securities at a later date. The repurchase price is greater than the original sale price. The difference between the two is the economic equivalent of interest, sometimes called the repo rate. Sometimes the sale and repurchase take place in very short order, often with only a day or two separating them. The scenario under contemplation, therefore, is that Northern Rock sells securities to the Bank of England and agrees to buy them back. In such a case what is the financial assistance that the Bank of England is giving to Northern Rock? Surely it is the agreement to buy securities from Northern Rock in the first place in return for Northern Rock’s promise to repurchase at a higher price. When that form of financial assistance is withdrawn all that it is necessary to assume is that the Bank stops buying securities from Northern Rock. I do not consider, therefore, that the repo argument takes the case further.

As I think Mr Howard accepted there is no express assumption in section 5 (4) (a) that Northern Rock had realised assets which in fact it had not. But he said that an assumption that Northern Rock had realised assets to repay the loans was the inevitable consequence of the assumption that the loans had been repaid. And that in turn was the true meaning of the assumption that the financial assistance had been withdrawn. In my judgment there are four problems with this submission. First, it assumes the end result that Mr Howard wants to assume and works backwards from that to the words of the statute. Second, it relies heavily on a series of counter-factual consequential assumptions that are not to be found within section 5 (4) (a). Third, there were a variety of ways in which Northern Rock could have realised assets, depending on the mix of cash, liquid securities and mortgages that it chose to realise. The valuer has made one series of assumptions. But why that series of assumptions, rather than another? Any set of assumptions about which assets Northern Rock is treated as having realised is no more than an arguable consequence of the express assumptions. Fourth, it pays no attention to the reality principle.

The last point to make on the detailed wording of section 5 (4) (a) is that, to my mind at least, a “withdrawal” is the unilateral act of the party doing the withdrawing. Yet the valuer’s interpretation entails reciprocal or responsive action on the part of the counter-party, who is not mentioned at all in section 5 (4) (a). What is necessarily entailed in a withdrawal depends on what is being withdrawn. Mr Howard rightly submitted that if what is being withdrawn is money in a bank account the withdrawal takes place when the money leaves the account rather than when the cheque is written. But that is because what is being withdrawn is the money itself. By contrast, if you are on my land and I withdraw my permission for you to be there the withdrawal does not cause you to leave; it merely turns you into a trespasser. That is because what is withdrawn is the permission. If I withdraw my ship from charterers for non-payment of hire I have done no more than terminate the charterparty. In reality the ship remains under the control of the charterers; it is not instantly returned to owners. So I do not consider that inherent in the concept of withdrawal is the notion that all loans have been repaid instantly.

Mr Howard submitted that what Mr Phillips was trying to persuade us to accept was that Northern Rock should be valued on the assumption that repayment of the Bank of England’s loans would take place over a period of time, thus enabling Northern Rock to retain income from its mortgage book. I think it fair to say that Mr McKillop’s expert report, prepared on behalf of Harbinger, did come close to making this assumption. But as I have said there is a critical difference in the context of a valuation between events up to and including the valuation date and events thereafter. The basic principle is that the valuer takes reality as it was on the valuation date, subject to any departures required by the hypothesis under which he values, but he makes no assumptions about the future. His job is to determine what price would be paid for the asset in question on the valuation date itself. Whatever might happen after the valuation date has not happened either in the real world or in the hypothetical world. The purchaser will no doubt consider all the possibilities, but he would be foolish in the extreme if he were confidently to predict the future.

For the reasons I have given I cannot see that the valuer is required to assume that Northern Rock has realised assets which in fact it owned at the transfer time. The reality was that, at least according to Northern Rock’s balance sheet, it was the owner of assets worth £106 billion. I cannot see that the statutory hypothesis requires the valuer to ignore or alter that real fact. The valuer’s assumption itself postulates a hypothetical transaction (or series of transactions) for which neither the Act nor the scheme expressly provides. Nor do I consider that the words require an assumption that Northern Rock has repaid loans which, in the real world, remained outstanding. They are, to borrow Scott J’s phrase, allegedly consequential assumptions which follow from the only explicit assumption, viz. that financial assistance has been withdrawn by the making of a demand.

I do not think that the assumption required by section 5 (4) (b) alters this conclusion. That assumption is that:

“no financial assistance would in future be provided by the Bank of England or the Treasury to the deposit-taker in question (apart from ordinary market assistance offered by the Bank of England subject to its usual terms).”

In order to understand that assumption it is necessary to have recourse to the statutory definition of financial assistance. When that is read into section 5 (4) (b) it becomes apparent that what is to be assumed is that no further loan will be provided by the Bank of England; and no further repo transaction will be entered into by the Bank of England. I do not consider that section 5 (4) (b) compels the conclusion that section 5 (4) (a) necessarily requires the counter-factual assumptions that the valuer’s interpretation entails.

In short in my judgment the wording of section 5 (4) (a) is not clear enough to trump the reality principle so as to compel the series of counter-factual assumptions that the valuer made. It is, I think, this that ultimately causes me respectfully to disagree with Mummery and Beatson LJJ on the outcome of the appeal.

In my judgment, therefore, the first of the two interpretations advanced by Mr Phillips is correct. I add that the alternative interpretation (namely that the Bank of England loans have been discharged by the notional transfer of assets at book value) also offends the reality principle, and is not required by the statutory assumptions. I would reject that interpretation.

The Tribunal said:

“In the context of loans, an actual loan will, in most commercial banking situations, commonly be made by drawing-down a facility. It can be said that the granting of the facility is “financial assistance” but even if that is so, the actual draw-down on the facility, thereby creating a relationship of creditor and debtor, is itself provision of financial assistance. It seems to us, therefore, that when section 5(4)(a) speaks of withdrawal of financial assistance it is speaking not only of withdrawing the facility but also of withdrawing the loan itself.” (§ 116)

“In the case of Northern Rock, it may have been that the facility afforded by the Bank had been fully drawn-down. But in other cases that may not be so. Withdrawing financial assistance may then encompass two elements, first the withdrawal of the facility so that further draw-down is prohibited and secondly, the withdrawal of the loan itself. We consider that the ordinary meaning of the word “withdraw” in relation to such a loan (and given that the word is, as we have said, used in a portmanteau phrase covering all types of financial assistance) is that the loan is no longer in place, in other words it has been satisfied. It would be a fallacy, in our view, to focus on the words “financial assistance” without regard to the nature of the particular financial assistance concerned. In the case of loan, it can, we accept, sensibly be argued that “assistance” is withdrawn when a demand for payment is made. But when one acknowledges that “financial assistance” is referring to particular types of support (eg a guarantee or a loan) the question is what is meant by withdrawing that support ie the guarantee or the loan. To withdraw a loan is not the same thing as to demand repayment of a loan. Indeed, the support or financial assistance which a loan gives in practice is not altogether withdrawn by a demand since the debtor still has the use of the money until it is repaid. If one were to ask whether a debtor had the benefit of a loan following a demand for payment but before actual repayment, the answer would clearly be Yes, in our view: financial assistance continues and it has not been “withdrawn” in the context of the Withdrawal Assumption.” (§ 117)

“We have considered why the draftsman considered that it was helpful to insert the words in parenthesis. Harbinger’s contention would suggest that part of the purpose of the words was to make clear that, in the case of a loan, the making of a demand was sufficient on its own to count as a withdrawal, and thus that actual repayment was not required. We cannot accept this. If this had been the draftsman’s intent, much clearer words would have been required. We have some doubt as to the draftsman’s actual purpose. We think the most likely explanation is that the words in parenthesis were included as a precaution. He was seeking to make clear that the assistance had to be assumed to be withdrawn whatever its nature or terms. So, for example, in the case of an on-demand loan, the method of withdrawal would be the making of a demand, so as to obtain repayment. But if (contrary, perhaps, to general expectations) some form of assistance had been given which could not be terminated immediately, the assistance must still be treated as withdrawn, even though in the real world the lender might not be able to withdraw it immediately.” (§ 119)

I agree with the Tribunal that the draftsman was seeking to make clear that the assistance had to be assumed to be withdrawn whatever its nature or terms. The Tribunal went on to say, correctly in my judgment, that “in the case of an on-demand loan, the method of withdrawal would be the making of a demand, so as to obtain repayment”. But if the method of withdrawal of that form of financial assistance is the making of a demand, then surely it follows that once the method of withdrawal has taken place the financial assistance has indeed been withdrawn. That to my mind is the full extent of the statutory assumption that is required to be made in the case of an on-demand loan. It seems to me, therefore, that there is an inconsistency in the Tribunal’s reasoning. On the one hand they say (correctly in my judgment) that the method of withdrawing financial assistance is by the making of a demand for payment. On the other hand they say that withdrawal of the loan does not take place until the loan is repaid. The latter involves an additional assumption to that which the Tribunal held was the method of withdrawing financial assistance, and which is the only (relevant) assumption laid down by the Act and the scheme.

I am not persuaded by the distinction between the “facility” and the “loan” which the Tribunal drew in [116]. Nor do I consider that the Act requires the assumption that the loan is withdrawn (whatever than means). Section 5 (4) (a) requires the assumption that the assistance is withdrawn. Section 15 (the definition section) defines assistance as including:

“assistance provided by way of loan…”

This definition continues to focus attention on the assistance. It does not transfer attention to the monies actually lent. While the Tribunal may be correct in saying that Northern Rock would continue to have the benefit of the money until it had been repaid, I do not agree that a debtor’s refusal to pay following a demand by a creditor amounts to the continued provision of financial assistance by the creditor. In my judgment the Tribunal were right in their conclusion that that the method of withdrawing financial assistance is by the making of a demand for payment; but wrong to make any further counter-factual assumption beyond a demand for repayment.

The Tribunal also considered other forms of financial assistance. We are not directly concerned with them, although I agree that an interpretation of the assumption ought to make sense across the whole range of potential financial assistance (as the Tribunal said at § 112). In the case of financial assistance which could not be terminated immediately, I agree with the Tribunal that the counter-factual assumption that all financial assistance had been withdrawn compels the conclusion that such financial assistance could be immediately terminated. In other words it becomes immediately repayable. The question of a guarantee is more problematic; but it does not arise for decision. As I read the Tribunal’s decision (§ 113) they took the view that the withdrawal of a guarantee had the consequence that it ceased to have effect both for the future and for the past. Harbinger does not challenge that conclusion. But Mr Phillips argues, and I agree, that the means by which financial assistance can be withdrawn must depend on the kind of financial assistance that is under consideration. I agree with him that it is not possible to extrapolate from one form of financial assistance to another.

To some (albeit a limited) extent I think that my interpretation of the assumption in section 5 (4) (a) is supported by the other assumptions required to be made. The support is limited because as Mr Howard correctly pointed out the assumptions required to be made by section 5 (4) (a) are mandatory for all schemes made under the 2008 Act, whereas other assumptions are, in his phrase, “optional extras.” The correct interpretation of section 5 (4) (a) should not change radically depending on whether the optional extras are or are not included in any particular scheme. Nevertheless, since entitlement to compensation depends on the scheme, the remaining provisions of the scheme cannot be totally ignored.

It will be recalled that the scheme requires the assumptions that Northern Rock:

“(a)

is unable to continue as a going concern; and

(b)

is in administration.”

Administration is governed by Schedule B1 to the Insolvency Act 1986. Paragraph 3 (1) of that schedule provides:

“The administrator of a company must perform his functions with the objective of –

(a)

rescuing the company as a going concern, or

(b)

achieving a better result for the company’s creditors as a whole than would be likely if the company were wound up (without first being in administration), or

(c)

realising property in order to make a distribution to one or more secured or preferential creditors.”

Paragraph 3 (3) provides:

“The administrator must perform his functions with the objective specified in sub-paragraph (1)(a) unless he thinks either:

(a)

that it is not reasonably practicable to achieve that objective, or

(b)

that the objective specified in sub-paragraph (1) (b) would achieve a better result for the company’s creditors as a whole.”

In our case the scheme requires it to be assumed that Northern Rock is unable to continue as a going concern. It necessarily follows that for the purpose of the valuation it must be assumed that the objective specified in paragraph 3 (1) (a) of Schedule B1 cannot be achieved, whether or not that is the case in reality. Accordingly the valuation must take place on the assumption that the purpose of the administration is to achieve a better result for the company’s creditors as a whole than would be likely if the company were wound up. In reality the Bank of England was one of the company’s creditors. I do not find in the Act or the scheme a clear assumption that the Bank of England has ceased to be one of Northern Rock’s creditors. In my judgment therefore the effect of these assumptions is that the administration is to be assumed to be conducted with the objective of securing a better realisation for creditors (including the Bank of England).

I have not found it necessary to deal with the argument about compatibility with Article 1 of the First Protocol to the European Convention on Human Rights. On my interpretation of section 5 (4) (a) it does not arise.

I might add that the interpretation I favour is consistent with the decision of this court in SRM Global Masterfund LP v HM Treasury [2009] EWCA Civ 788 which was the original unsuccessful challenge by judicial review to the basis of compensation. It should be borne in mind that the challenge alleged that Northern Rock should have been valued on the basis that it was a going concern and was not in administration. The point raised in this appeal was therefore not explicitly considered in that case. But the judgment of Laws LJ contains the following passages:

“Yet, so the argument goes, that is the result which the s.5(4) assumptions (and, for good measure, the assumptions required by paragraph 6 of the Compensation Scheme Order) produce. Their consequence is that the shares in the company fall to be valued for the purposes of compensating the deprived shareholders on the basis of a "fire sale" – a forced sale of the assets of a company in liquidation (or other form of insolvency procedure); circumstances in which, notoriously, the return obtained is very depressed.” (§ 39) (Emphasis added)

“Yet in the result the shareholders are altogether stripped of the assets' value, save for whatever net sum a fire sale in the course of a liquidation or other administrative procedure might bring: which is likely to be nothing.” (§ 63) (Emphasis added)

“Yet more obvious is the linked proposition that there will be some residual value in the assets if on a fire sale what is got for them exceeds the company's liabilities (the administrator or liquidator being obliged to obtain the best price).” (§ 76) (Emphasis added)

Although the point raised by this appeal was not in issue, it is noticeable that the case was argued on all sides on the footing that sales would take place in the course of the administration or liquidation and not before. Whether the sale would or would not be a “fire sale” is not a matter of legal interpretation of the assumption. It is a question of what the purchaser would perceive.

That is not, of course, to say that the shares are necessarily valuable. Given that there was no buyer for Northern Rock in the real world (absent continuing State support), coupled with the Bank of England’s policy of recovery and the background of EU law, they may well have been worthless. But in ascribing a value to the shares the valuer must adopt the correct legal framework.

Finally, in the course of the hearing we raised the question whether the Tribunal had made an order giving effect to its decision. Normally, appeals are brought against orders rather than reasons for decisions. However, on further investigation it seems that that does not apply to the Upper Tribunal. Rule 40 of the Tribunal Procedure (Upper Tribunal) Rules 2008 requires the Upper Tribunal to provide “a decision notice” and “written reasons for its decision”. Section 13 of the Tribunals and Enforcement Act 2007 confers a right of appeal to the Court of Appeal “on any point of law arising from a decision made by the Upper Tribunal”. That right is circumscribed by article 2 of the Appeals from the Upper Tribunal to the Court of Appeal Order 2008 (which says that there must be an important point of principle etc), but does not otherwise affect this question. Thus the right of appeal is not drafted by reference to an order, but by reference to a decision. It appears, therefore, that the decision notice takes the place of an order, and that no order is necessary before a right of appeal arises. Accordingly, there was no need for concern about the absence of an order.

Left to myself, therefore, I would allow the appeal on the question of interpretation; and remit the case to the valuer for reconsideration. However, since Mummery and Beatson LJJ disagree with my interpretation, the appeal must be dismissed.

The crucial question in this appeal is the meaning of the statutory assumption required by section 5(4)(a) of the Banking (Special Provisions) Act 2008 (“the 2008 Act”) for the purposes of determining compensation, in this case to the shareholders of Northern Rock plc when it was nationalised. Section 5(4) provides that it “must be assumed”, “(a) that all financial assistance provided by the Bank of England or the Treasury has been withdrawn (whether by the making of a demand for repayment or otherwise)…”. Is what “must” be assumed that “all financial assistance” has been “repaid”, or is what must be assumed that the Bank of England or the Treasury have made a demand for repayment? The two interpretations have respectively been characterised as the “Repayment Interpretation” and the “Demand Interpretation”. As the Appellants’ expert reports modelled a five year run-off in administration (the First Respondent’s consultation document was similar), the Demand Interpretation might have been characterised as “the Administration interpretation”. It is not necessary for me to deal with a third interpretation, the “Repayment-in-Kind Interpretation” (explained by Mummery LJ), which was the Appellants’ primary case before the Tribunal, and their secondary case in their written submissions for this court.

The First Respondent adopted and applied the “Repayment Interpretation”, and assumed that Northern Rock would have had to accept a 15% discount to sell enough of the mortgages it held to enable it to repay the loans made to it by the Bank of England forthwith. A balance sheet surplus of over £1.6 billion became a deficiency of £2.4 billion. The Appellants maintain that he and the Tribunal erred in applying that interpretation. They contend that, had the “Demand Interpretation” been adopted the shares would have had greater value. As Lewison LJ observed (at [65]), this may or may not be the case, but the value of the shares is not the question before this Court. We are concerned only with whether the Tribunal (and the First Respondent) erred in adopting the Repayment Interpretation in order to ascribe a value to the shares.

I gratefully adopt the account of the facts and the summary of the legislation by Lewison LJ, but, while respecting the force of his analysis as to the meaning of section 5(4)(a), I respectfully disagree with his conclusion rejecting the “Repayment Interpretation”. I agree with Mummery LJ that the 2008 Act requires it to be assumed that all financial assistance has been repaid and with his approach to and conclusion about the “reality principle”. As we are divided, I add a few words of my own on the meaning of the assumption in section 5(4)(a). I also explain why I do not consider that the “Repayment Interpretation” is incompatible with Article 1 of Protocol 1 to the European Convention on Human Rights and refer briefly to the status of the point raised by the Appellants as to the scope of the Tribunal’s jurisdiction.

The background to section 5(4) and its context was a general agreement that the “Lender of Last Resort” principles applied and that the government should not have to compensate the shareholders of Northern Rock for value which was dependent on taxpayers’ support. In the light of those considerations, I do not consider that it can be said that the financial assistance provided by the Bank of England or the Treasury can be said to have been “withdrawn” by a demand for repayment. My principal reason is that, as a matter of ordinary language, in the case of a loan, a demand for repayment does not itself withdraw the financial assistance. The assistance will continue until the loan is repaid. The debtor, here Northern Rock, will, as the Tribunal stated (at [117]) have the benefit and the use of the money until the creditor is repaid.

Where the assistance has taken the form of a guarantee or a repo agreement, such as the Bank of England repo agreement over Northern Rock’s Treasury Investments, the mechanism of withdrawal will vary, but the statutory assumption is that the assistance has ended, either by the guarantor cancelling the guarantee so it ceases to have effect or the Bank of England selling the securities in a repo agreement. I agree with Mummery LJ that the assumption required by the statute is that the relevant process has ended rather than that it has only begun.

The “Demand Interpretation” is also difficult to apply consistently to different forms of financial assistance. Adopting it would mean that the value of Northern Rock’s shares depends on the form of financial assistance. In the case of a repo agreement, where the Bank of England is able to sell the collateral immediately, the result is the same as under the Repayment Interpretation. In the case of a loan for a fixed term, the result could be very different because, absent a default, the demand would have no effect.

What is the impact of the words in parenthesis at the end of section 5(4)(a)? Lewison LJ has stated (at [40]) that the parenthesis was “intended to illustrate how financial assistance is to be withdrawn”, that “one way is by making a demand for repayment”, and “in other words, the making of a demand is the means of withdrawing financial support”. He concluded (at [41]) that a demand “for repayment looks forward to a future repayment … but it does not necessarily entail an immediate repayment”. I respectfully disagree. I consider that the Tribunal was correct (at [118]) in reading the words as referring to mechanisms for withdrawal, and accept Mr Howard’s submission that the words “or otherwise” show that the method of withdrawal does not matter and that what matters is the effect of the withdrawal of financial assistance.

The consequence of regarding one method of effecting a withdrawal of financial assistance as “making a demand for repayment” so that the statutory assumption would be met by the demand is that Northern Rock (in reality its administrator) could spread the repayment of loans over an undefined period of months or years without being regarded in the interim as in receipt of financial assistance by the Bank of England or the Treasury. This would be so although Northern Rock would have the use of the loans or part of them until such time as they are fully repaid. I do not consider that to be compatible with the principle underlying the compensation provisions in the 2008 Act; that the government should not be required to compensate the shareholders of Northern Rock for value which was dependent on the support that was given by taxpayers. I agree with the Tribunal (at [130]) that “only the Repayment Interpretation fully satisfies the aim of ensuring that shareholders do not reap the benefit of taxpayer support and of the taxpayer bearing the inherent risk in that support”.

The Appellants submitted that the Repayment Interpretation is not compatible with Article 1 of Protocol 1 to the European Convention on Human Rights (hereafter “A1P1”) and that section 3 (or alternatively section 6) of the Human Rights Act 1998 requires the adoption of the Demand Interpretation which is compatible with it. A1P1 provides:

“Every natural or legal person is entitled to the peaceful enjoyment of his possessions. No one shall be deprived of his possessions except in the public interest and subject to the conditions provided for by law and by the general principles of international law. The preceding provisions shall not, however, in any way impair the right of a State to enforce such laws as it deems necessary to control the use of property in accordance with the general interest or to secure the payment of taxes or other contributions or penalties.”

The basis upon which compensation for property which has been nationalised is calculated is an important factor in determining whether the “taking” of the property respects a fair balance between the interests of the former property holders and those of the government and the wider public, and whether the “taking” imposes a disproportionate burden on those former property holders: James v United Kingdom (1986) 8 EHRR 123 at [54];Lithgow v United Kingdom (1986) 8 E.H.R.R. 329; Holy Monasteries v Greece (1994) 20 EHRR 1 at [71]. But, as was stated in the Lithgow and Holy Monasteries cases, A1P1 does not guarantee a right to full compensation in all circumstances.

In the context of these proceedings, what is of particular significance is that it is well established in the Strasbourg jurisprudence that a state enjoys a wide margin of appreciation as to how the compensation for a “taking” is calculated. Moreover, the judgment of the national legislature will be respected unless the assessment is “manifestly without reasonable foundation”: Lithgow v United Kingdom (1986) 8 E.H.R.R. 329at [122].

Although the Appellants submitted that the Repayment Interpretation is not compatible with A1P1, they accepted that, if contrary to their case, the First Respondent applied the correct assumption when valuing the shares, there are no grounds for overturning the nil valuation. But, if that is so, it follows, in the light of the cases to which I have referred, that the Appellants’ submission that this is incompatible with A1P1 is unsustainable. First, on this hypothesis, that would be the correct value of the shares at that time. Secondly, irrespective of whether it was in fact the correct valuation, it cannot be manifestly unreasonable to leave out of account such value in the shares as accrued from having the benefit of government funds during an administration which could have lasted for a considerable period.

In the light of my conclusion, it is not necessary to deal with the Appellants’ contention that the Tribunal fell into error because its decision on this question was primarily based on this Court’s decision in R (SRM Global Master Fund LP) v Treasury Commissioners [2009] EWCA Civ 788, [2010] B.C.C. 558. The Appellants are correct in stating that Laws LJ’s judgment in that case did not address the interpretations canvassed in these proceedings, and did not consider a “fire sale” before nationalisation, only such a sale during administration, and that the evidence before this Court differs from that in the judicial review proceedings. Those differences are not of importance in assessing whether Laws LJ’s judgment is of assistance in these proceedings. In my judgment it is. In considering the compatibility of section 5(4) with A1P1, Laws LJ was envisaging the possibility of a valuation based on a notional “fire-sale” of assets which produced a nil value. His conclusion was that, notwithstanding such nil value, there was no incompatibility because (see [76]) “the business is shown to be worthless without the support put in by the government” is based on and reflects the Strasbourg jurisprudence to which I have referred and not the differences on which the Appellants relied.

Finally, I turn to what was referred to as “the section 133(5) question”. Section 133(5) of the Financial Services and Markets Act 2000 as amended by paragraph 18 of the Compensation Scheme Order 2008 provides that when the Tribunal concludes that a decision as to the amount of compensation is not “a reasonable decision”, it must remit the matter to the Valuer for reconsideration”. Section 9(6) of the 2008 Act inter alia states that the provision that may be made under that Act for a tribunal does not include enabling the Tribunal “to substitute its own decision for that of the person who made the decision”.

Mr Pushpinder Sani submitted that, as the First Respondent’s valuation was an administrative decision and not one by a judicial tribunal, and as the Appellants’ private law rights depended upon it, Article 6 of the ECHR meant that the Tribunal had “full jurisdiction” to consider all arguments of fact and law going to the merits of the valuation and erred in not doing so. He maintained that the interpretive direction in section 3 of the Human Rights Act 1998 enabled the Court to read section 9(6) of the 2008 Act compatibly with Article 6. These questions did not (see Tribunal’s decision, [266] – [267]) arise in the Tribunal in view of its decision on the assumption required by section 5(4)(a) of the 2008 Act; that is giving the Tribunal full jurisdiction. In view of the conclusion reached by Mummery LJ and myself, they also do not fall for decision in this appeal. In those circumstances, I confine myself to two observations. First, the power under section 3 is very broad. It is not bounded by the constraints of linguistic possibility, but only by the limits of constitutional propriety. However, before getting to the section 3 question it is necessary to conclude that there is an incompatibility between a legislative provision and the ECHR. My second observation is that in these proceedings, the submission that the totality of the procedure before the First Respondent and the Tribunal does not comply with Article 6 faces a number of significant hurdles.

For the reasons I have given, as well as those in Mummery LJ’s judgment, I would dismiss this appeal.

According to the independent valuation produced by Mr Andrew Caldwell (the Valuer) in his revised assessment notice dated 1 October 2010 no compensation is payable by HM Treasury (the Treasury) to the Northern Rock shareholders for the shares taken into “temporary” public ownership on 22 February 2008 (the valuation date). The Upper Tribunal found no error of law in his valuation of the shares at zero, nor can I. Harbinger disagrees. It says that, in determining the amount of compensation payable by the Treasury for the shares, the Valuer misinterpreted the “Withdrawal Assumption” in s.5(4)(a) of the 2008 Act. The critical question for the court is: what is the natural and ordinary meaning of the Withdrawal Assumption?

The Upper Tribunal gave detailed consideration to three different interpretations before it came down in favour of the “Repayment Interpretation” adopted by the Valuer. It rejected the “Demand Interpretation” and the “Repayment-in-Kind Interpretation” advanced by Harbinger. The judgment of the Upper Tribunal and the submissions to this court used those labels for convenience. I will use them in this judgment, but they are only labels. The concepts to which they have been attached may be more complex than the labels.

According to the Repayment Interpretation the Withdrawal Assumption involves a valuation assumption that the financial assistance of billions of pounds provided to and received by Northern Rock from the Bank of England (the Bank) and the Treasury was itself immediately removed on the valuation date: the assistance was discharged and satisfied by notional repayment on that date, leaving no financial assistance to Northern Rock outstanding after that date.

The Demand Interpretation focuses on the notional making, on the valuation date, of a demand for repayment, regardless of the date on which the financial assistance would be notionally repaid.

Against that interpretation it is said that a mere demand would be insufficient in itself to amount to financial assistance being withdrawn. It would change nothing. It would have no effect on, or relevance to, the value of the shares as at the valuation date. A mere demand would leave Northern Rock with the continuing use and benefit of financial assistance until repayment itself had subsequently taken place.

Further, that approach to valuation cannot be squared with the intended purpose of the Withdrawal Assumption: it would involve having to ascribe some value to the shares in consequence of Northern Rock’s continuing retention of the use and benefit of the financial assistance received by it from the Bank and from the Treasury. The notional withdrawal of that assistance would become a notional process spread out over a number of years, such as might occur in the course of an administration, rather than be a discrete notional event prescribed as the basis for putting a value on the shares as at the specified valuation date.

In support of the “Demand Interpretation” Harbinger calls attention to the context of a hypothetical administration rather than a liquidation of Northern Rock for the purposes of the share valuation exercise, though it does not rely on the “Administration Assumption” as itself affecting the interpretation of the Withdrawal Assumption. Harbinger emphasises the character of the financial assistance provided to Northern Rock as being “liquidity support” enabling it to pay its debts as and when they fell due. It points to the fact that, before financial assistance began to be provided on 13 September 2007, Northern Rock was balance sheet solvent with a surplus of about £2bn. That fact informed the Withdrawal Assumption in the subsequent emergency legislation. The consequence of adopting the Repayment Interpretation is that, as a matter of valuation, £4bn was removed from the Northern Rock balance sheet.

On the Demand Interpretation the provision of financial assistance as “liquidity support” would be treated as withdrawn on the making of the notional demand for repayment on the valuation date, regardless of whether or not the demand was met on that date. The notional demand would itself be a calling in and a discontinuance of the financial assistance which had been provided for the “liquidity support” purpose. The assistance would be withdrawn in the sense that it was no longer available, either in the form of guarantees or loan facilities. As a matter of commercial reality, the very making of the demand, without the necessity for immediate repayment, would inevitably lead to a state of cash flow insolvency; and that, as a matter of law, would be a situation subject to a statutory insolvency process, such as administration.

Nothing expressly stated in s.5(4) required an assumption of repayment itself on the valuation date or the immediate discharge or satisfaction of the financial assistance on that date. Harbinger objects that the assumption of repayment precipitated a situation in which there would be an instantaneous “fire sale” or a “distress sale” of Northern Rock’s assets with a consequent reduction in the value of Northern Rock and of its shares in the manner wrongly determined by the Valuer. The consequence of his acceptance of the Repayment Interpretation was that, in the hypothetical administration of Northern Rock, its opening balance sheet surplus of over £1.6bn became a deficiency of £2.4bn.

Harbinger’s “fire sale” or “distress sale” description of what would happen on the Repayment Interpretation is disputed by the Valuer. Its position is that, in making his nil valuation, the Valuer took nothing out of Northern Rock that would otherwise be there: the Valuer rightly recognised that the book value of Northern Rock’s assets, as at the time when the balance sheet was prepared, did not reflect their realisable value as at the valuation date.

According to the alternative “Repayment-in-Kind Interpretation” advanced by Harbinger, but also rejected by the Upper Tribunal, Northern Rock is treated as having repaid the financial assistance in kind by a transfer of assets at book value. There would be a “swapping out” of the financial assistance from Northern Rock’s balance sheet for an equivalent value of Northern Rock’s unencumbered mortgage assets immediately before the valuation date. The main difference between that interpretation and the Repayment Interpretation was that the latter treated the assistance as being repaid in cash, while the former assumed that assets at book value would be accepted in repayment of the assistance, regardless of their actual value at the valuation date.

According to Harbinger the consequence of both interpretations advanced by it is that the shareholders in Northern Rock would be in no better position than they would have been before the financial assistance was provided. That was consistent with the purpose of the Withdrawal Assumption. Harbinger contends that the Repayment Interpretation adopted by the Valuer is inconsistent with that purpose, because it puts the shareholders of Northern Rock in a worse position than they were in before the provision of financial assistance in September 2007.

The significance of that date is that it was when the auditors reported that they had reasonable grounds for believing that Northern Rock would be unable to continue as a going concern. That report was soon followed by media report that Northern Rock would be seeking financial assistance and by a run on it for withdrawal of deposits. The Bank then provided financial assistance, in its role as lender of last resort, to prevent the collapse of Northern Rock and other parts of the financial sector. In addition to the secured loan made by the Bank (£27bn) the Treasury had guaranteed all retail deposits with Northern Rock (£29bn).

It is worth noting at the outset the role of the Upper Tribunal and of this court: neither is a valuation body that makes primary findings of fact about how much the shares were worth at the valuation date. Further, these are not judicial review proceedings, such as were brought by the Northern Rock shareholders in SRM Global Masterfund LP v. HM Treasury [2009] EWCA Civ 788 when mounting a legal challenge to the statutory basis prescribed for the assessment of compensation as incompatible with the right to property under Article 1 of the First Protocol to the European Convention on Human Rights (ECHR). It was contended in those proceedings that the inevitable effect of the valuation provisions was that the Valuer was bound to find a nil value for the shares. That is not Harbinger’s present position.

Nor are these insolvency proceedings, such as might have been brought, if the Bank and the Treasury had not come to the rescue of Northern Rock with the package of emergency measures for financial assistance that saved it from being put into administration or liquidation.

This appeal is confined to a point of statutory interpretation. As already indicated, it turns solely on whether the independent expert valuation of the Northern Rock shares made by the Valuer was vitiated by an error of law. Did the Valuer act on an assumption that was different from what he was required by the legislation to make about the withdrawal of financial assistance? This court has to decide whether the Valuer and the Upper Tribunal correctly interpreted what s. 5(4) means. If there was an error in the interpretation of the Withdrawal Assumption, this court should remit the matter for the Valuer to make a fresh valuation based on the interpretation decided upon by this court.

For the reasons given in his judgment, which I have read in draft, Lewison LJ says that this court should take that course. I adopt his concise account of the facts and his summary of the legislation, but, with great respect, I am unable to agree with his carefully reasoned conclusion that the Valuer should be directed to re-visit his valuation on the basis of a different interpretation of s.5(4) than that adopted by him first time round.

The valuation exercise is not controversial in most respects.

The Valuer’s function was to place a monetary value on the shares in Northern Rock. He was to do that as at the date of their compulsory acquisition by the Treasury. The duty of the Valuer was to value the shares in accordance with the Withdrawal Assumption in s.5(4)(a) of the emergency legislation. He was not free to question, qualify or contradict the Withdrawal Assumption, for example by reference to evidence of real facts that might affect the value of the shares. Such evidence would not be relevant to a valuation exercise based on a departure from the real world required by the statutory hypothesis. Real facts would only be relevant on matters that were not required to be assumed by the Withdrawal Assumption.

The immediate background to the shares being taken into public ownership and to the Withdrawal Assumption basis of valuation was that no private sector buyer could be found who was prepared to buy the shares in Northern Rock, unless the Treasury and the Bank continued to provide financial assistance to Northern Rock. The legislation posited the Withdrawal Assumption as the basis on which the shares were to be valued as at the valuation date. The aim of the statutory hypothesis was obvious: it was to ensure that the element of share value attributable to the package of measures of financial assistance provided, or to be provided, in various forms at the expense of the taxpayer (via the Bank and the Treasury) should be completely eliminated from the assessment of the amount of compensation, particularly as that compensation would be at the ultimate expense of the taxpayer (via the Treasury) to the shareholders in Northern Rock.

In assessing the compensation at nil the Valuer proceeded on the basis that he was required to assume, contrary to the real facts, that, on the valuation date Northern Rock had (a) repaid all the monies lent to it by the Bank; and (b) realised sufficient of its assets to enable it to make that repayment: see para 2.30 of the Consultation Document in which the Valuer explained his assumptions about how the withdrawal of financial assistance referred to in s.5(4)(a) was achieved. The Valuer’s interpretation and application of the Withdrawal Assumption resulted in a marked discount in the book value of the majority of Northern Rock’s assets.

That approach was upheld by the Upper Tribunal. It is now challenged as a matter of law. The main point made is that in September 2007 Northern Rock had liquidity problems: it was unable to meet its debts as and when they fell due. It needed “liquidity support” facilities. They were provided by way of financial assistance in the form of guarantees, secured loans and “repo” facilities. However, according to its balance sheet, Northern Rock was solvent: the value of its assets exceeded the amount of its liabilities by a substantial surplus of £1.6bn in February 2008. The Valuer’s valuation of the shares at nil took no account of that fact. Instead, his valuation was made on the basis of converting the balance sheet surplus into a deficit position of £2.44bn. That reduced the value of Northern Rock by £4bn. The hypothetical sale of the shares at a discount to reflect a “fire sale” or a “distress sale” thus eliminated the balance sheet surplus from the valuation exercise.

The error in his interpretation and application of the Withdrawal Assumption is, Harbinger says, that the Valuer ignored relevant real facts in arriving at his valuation. That is as far as the challenge goes, for Harbinger accepts that if, contrary to its case, the Valuer followed the correct approach, there are no grounds for overturning the nil valuation.

What is the true ambit of s.5(4)? The background is touched on above and is set out in detail in the judgment of Lewison LJ. One aspect of it is that the financial assistance, which the Valuer was directed to assume was withdrawn on the valuation date, was by way of a package of measures including guarantees, “repo” facilities and loans of very large sums of money repayable on demand. The precise point of disagreement is whether the Withdrawal Assumption involved only the making of a notional demand on the relevant date for repayment of the assistance to be made, as Harbinger contends, or whether it involved making notional repayment of the assistance, as the Treasury contends and as was accepted by the Valuer and upheld by the Upper Tribunal.

I turn to the wording of s.5(4). Taken with the inclusive description of “financial assistance” in s. 15(1), the clearly stated assumption is that on the valuation date all financial assistance “has been withdrawn.” That wording is wide enough to cover the withdrawal of all kinds of financial assistance provided by the Bank or the Treasury to Northern Rock. Under the description in s. 15(1) different forms of financial assistance provided are expressly “included”, such as assistance by way of loan, guarantee and indemnity, as well as “repos” (transactions involving the sale and purchase of securities) and is plainly wide enough to include money repayable on withdrawal of financial assistance.

The Withdrawal Assumption covers financial assistance provided by the Bank and the Treasury to Northern Rock, whether before or after the passing of the Act: the required assumption is that all financial assistance provided “has been withdrawn” and that no assistance “would in future be provided.” That covers money paid over in the past, as well as money that has, in the past, been promised to be paid as assistance in the future.

The wording of the Withdrawal Assumption does not indicate a notional process that would take place in the real world over a period of time. Rather, it indicates a state of affairs that must be assumed to have been in existence, for the purposes of valuation, as at a particular point in time (the valuation date). I see nothing in the wording of s.5(4) which would justify interpreting the Withdrawal Assumption as implementing a particular notional procedure or process for withdrawal of assistance, such as the making of a prior demand for repayment to be followed by a period for recovery of the assistance. What matters for the purposes of the hypothesis is that the financial assistance “has been withdrawn” and that all of that is assumed to have happened at a particular date, not that the assistance has been withdrawn in a particular way.

The expression “financial assistance” is itself a compendious description of the package of financial measures put together to support Northern Rock. The specific transactions or arrangements, whether they be loans, guarantees or simple money payments, were particular legal mechanisms used as a means of providing the financial assistance. The particular means or procedure appropriate for use in the actual withdrawal of financial assistance might in fact differ from case to case. Thus, in the case of a guarantee, the assistance would be withdrawn by the guarantor cancelling the guarantee. In the case of an agreement to lend money in the future, the assistance would be withdrawn on termination of the contract. In the case of assistance in the form of payment of money a demand for repayment would normally be made as the initial step towards obtaining repayment of the money. The demand would be just the start of a process of withdrawing financial assistance, which may take time to complete and only be completed on repayment by the person to whom it was provided.

The Withdrawal Assumption is that, on the relevant date, the financial assistance has, for valuation purposes, been completely removed; all over, finished, at an end. It is deemed not to exist any more as a relevant factor potentially affecting the value of the shares. In my view, an assumption in those comprehensive hypothetical terms refers to the assumed repayment on the valuation date of the money provided as financial support, not the making of a prior demand as the initial step to securing repayment in fact. Assistance is not assumed to have been withdrawn by a notional demand initiating a process of withdrawal.

It is true that s. 5(4) does not refer in terms to assistance being “repaid”, or to “repayment” of the assistance, save in the parenthesis (see below), or to “recovery” of the financial assistance There is an explanation for that: those particular expressions are not appropriate descriptions of all the different kinds of measures making up the financial package, whereas “withdrawn” is a broad term capable of covering all the different ways in which the financial assistance may be assumed to come to an end.

In brief, in the case of financial assistance in the form of money paid to Northern Rock, the ordinary and natural meaning of the Withdrawal Assumption that the financial assistance “has been withdrawn” refers to an assumed repayment on the valuation date, not just to deemed demand on that date for repayment of the money taking place at a future date. It is inherent in the assumption that that form of financial assistance “has been withdrawn” that there is assumed to be instant repayment of the assistance on the valuation date.

That result is consistent with the long-established approach to interpreting a statutory hypothesis. Assumed repayment of assistance on the valuation date may be properly described as an inevitableconsequence of the hypothesis that, on 22 February 2008, all financial assistance “has been withdrawn.” As was famously said in East End Dwellings Co Ltd v. Finsbury BC [1952] AC 109,132-133:-

“If you are bidden to treat an imaginary state of affairs as real, you must surely, unless prohibited from doing so, also imagine as real the consequences and incidents which, if the putative state of affairs had in fact existed, must inevitably have flowed from or accompanied it…The statute says that you must imagine a certain state of affairs, it does not say that having done so, you must cause or permit your imagination to boggle when it comes to the inevitable corollaries of that state of affairs.”

Lord Asquith’s elegant dictum has been repeated and applied across a wide range of hypothetical facts, assumptions, and other statutory deemings which are contrary to the facts in the real world. The consequences inevitably entailed in the Withdrawal Assumption are also to be assumed for valuation purposes and they may also be consequences that are contrary to the circumstances in the real world.

The parties disagree about what consequences are inevitably entailed in the Withdrawal Assumption. I take the view that, on its ordinary and natural meaning, the Withdrawal Assumption necessarily entails acceptance of the consequence that Northern Rock must be treated as having repaid all the financial assistance on the valuation date. Only if the assistance is treated as having been repaid on that date would Northern Rock cease to be in the state of receiving the benefit and use of financial assistance. A demand for repayment does not itself terminate the assistance. It is only a prelude to the ending of financial assistance. If the assistance is not treated as repaid by Northern Rock at the valuation date, the provision of financial assistance to Northern Rock, in the form of outstanding loans that have not been repaid, is being treated as continuing after the valuation date, as it was before. That is contrary to the intended instantaneous impact of the Withdrawal Assumption. If not treated as instantly repaid that financial assistance would be treated as continuing into the future and for as long as repayment of the loans to Northern Rock was outstanding. That financial assistance extending into the future would be at variance with the required assumption of no provision of financial assistance in the future in s.5(4)(b).

The assumed consequence of repayment of loans also necessarily entails the further consequence that Northern Rock must be treated as having had to realise all its assets on the valuation date in order to make the assumed repayment of the outstanding loans on that date. Hence the Valuer’s discounted value of the assets shown in Northern Rock’s balance sheet.

The fact that s. 5(4)(a) assumes that a demand has been made for repayment of the outstanding assistance says nothing and changes nothing about the provision of financial assistance. The financial assistance that has been provided is enjoyed as financial assistance, unless and until the assistance is fully repaid. I understand Harbinger’s point is that the assistance was in the form of “liquidity support” and that that form of assistance is withdrawn as soon as a demand is made for repayment, because funds are no longer available to pay debts as they fall due. However, that does not meet the point of Northern Rock’s continuing benefit and use of the financial assistance received until that assistance has been repaid.

Other points are made in support of a different interpretation of s.5(4).

The reality principle explained and relied on by Lewison LJ did not feature in the judgment of the Upper Tribunal or in the skeleton arguments of the parties in that Tribunal or in this court. It surfaced when, prior to the hearing and with my agreement, Lewison LJ asked counsel to consider the relevance of the reality principle in the light of two authorities. Those cases were not cited to the Upper Tribunal, though I doubt whether they would have made any difference to the outcome of the case which, at the end of the day, turns on the interpretation of the ambit of the Withdrawal Assumptions required by the wording of s.5(4).

In this case my view is that the reality principle tells us no more about the Withdrawal Assumption than is gathered from its wording, as interpreted in accordance with established principles. The reality principle is only saying that departure from the real world must be no greater than is required by the statutory Withdrawal Assumption. If you are not required by statute to depart from reality, you must stick with reality. But the principle does not determine or limit what the statute commands us to assume contrary to reality. The statute determines that. The reality principle is about what is not covered by the statutory assumption. What does not have to be assumed is real and what is real can be the subject of evidence that may be relevant the application of the Withdrawal Assumption to the real facts. If, in consequence of the interpretation of s.5(4), a matter falls outside the ambit of the Withdrawal Assumption, then it is real world matter open to evidence and argument.

The key question remains that of the interpretation of the ambit of the Withdrawal Assumption. If it does not extend to the assumed consequences of repayment of the assistance and the realisation of assets for that purpose at the valuation date, then I would agree that the Valuer and the Tribunal wrongly took too broad a view of the statutory assumption and acted on the wrong basis.

Particular emphasis is placed by Harbinger on the use in s.5(4) of the word “withdrawn”, rather than words such as “repayment” or “repaid.” The point is made that s.5(4) refers to the “withdrawal” of financial assistance and, in the case of a loan, withdrawal of assistance is withdrawal of assent, such as by a demand for repayment, which is unilateral: it is also an act prior to and short of repayment. A loan is repaid in consequence of withdrawal of financial assistance rather than repayment in itself being a withdrawal of assistance. So Harbinger contends that the Valuer was not required by the Withdrawal Assumption to assume that the loans had actually been repaid at the valuation date, as distinct from being called in by demand. If they are not to be assumed to have been repaid, then it would be wrong to assume a sale of the assets at a discount. The assumptions by the Valuer and by the Tribunal were departures from reality that were not justified by the Withdrawal Assumption.

I am not persuaded by those arguments. “Withdrawn” is wide enough, in the context of the provision of financial assistance, to describe a process that has been completed: what must be assumed to have happened on the valuation date is a totally completed process of withdrawal of financial assistance, not just the taking of a first step towards withdrawal by a process beginning with a demand.

The point on “withdrawn” ties in with the words in the parenthesis in s. 5(4)(a) “(whether by the making of a demand for repayment or otherwise)”, in particular the reference to “by the making of a demand for repayment.”

I agree with Mr Howard QC and the Upper Tribunal at [118] that, read in context the added words “or otherwise”, that parenthetical phrase does not invalidate the approach taken by the Valuer to the meaning of “withdrawn.” That concept is capable of a wider meaning than a demand for repayment. In the context of assistance being withdrawn by a person who has been providing that assistance, “withdrawn” covers the removal of various forms of financial assistance having different mechanisms for bringing them to an end. A guarantee which is cancelled is withdrawn in a different way than a loan, which is repaid.

The words inserted parenthetically clarify the point that the means or mechanics for the withdrawal of a particular form of assistance, such as by making demand for repayment of a loan, do not matter for valuation purposes. What matters for valuation purposes is that financial assistance provided in any form is treated as having been removed and come to an end on the valuation date for all purposes past, present or future. Repayment is treated as having been made on that date.

As I have already explained, Harbinger’s interpretations do not meet the objection that financial assistance would not be withdrawn by the making of a demand: withdrawal would only be complete on repayment of the assistance and notional repayment entails having to treat the assets of Northern Rock as sold on the valuation date for that purpose.

I agree with the Tribunal that Valuer’s Repayment Interpretation is compatible with the EU rules on state aid and Article 1 of the First Protocol to the Convention. It does not give rise to any illegality or absurdity as contended by Mr Phillips QC. I agree with the judgment of Beatson LJ on the ECHR point.

The above interpretation of s.5(4) is compatible with the additional assumptions required by paragraph 6 of the schedule to the 2008 Scheme Order, that Northern Rock is unable to continue as a going concern and that it is in administration.

The above interpretation of the assumption is unaffected by the fact that Northern Rock’s assets, as recorded in its balance sheet, exceeded its liabilities. That fact is only relevant if the assumptions are more narrowly construed so as not to require notional repayment on the valuation date and as not to involve the realisation of assets on that date for that purpose.

When, in a case like this, you are faced with detailed arguments on interpretation and they range widely, there is a risk of straying further and further from the few words on which the case really turns. It is reasonable to assume that when s.5(4) was drafted and debated in the midst of a serious crisis and under considerable pressure of time it was with nothing like the amount of energy and ingenuity that have been invested by the parties in their efforts to persuade the Upper Tribunal and this court what it means. There are risks in over-interpretation.

Of course, I recognise that there are also risks in over-simplification. The simple truth is that there are limits to the process of rationalising an interpretation, and you find that you have reached, as I have, the point of seeming to say the same thing over and over again in different words. For the reasons given above at length I think that the Valuer got it right in his short statement of his approach to his task.

I would dismiss the appeal.

Having considered the extent of the departure from reality required by the terms of the Withdrawal Assumption ([108] onwards) the Tribunal correctly concluded at [168] that the Repayment Interpretation relied on by the Valuer was to be preferred to the Demand Interpretation and the Repayment-in-Kind Interpretation advanced by Harbinger.