UKCloud Ltd (in liquidation), Re

Gareth Jonathan Allen, in his capacity as Official Receiver and liquidator of UKCloud Ltd, seeks directions pursuant to s 168(3) Insolvency Act 1986 as to whether the effect of a debenture granted by the Company to Harbert European Speciality Lending Company II SARL on 28 February 2020 is to give Harbert the benefit of a fixed or floating charge over certain internet protocol (IP) addresses. Harbert’s position is that it has a fixed charge. The OR is professedly neutral, but, whilst recognising that there is some force in Harbert’s arguments and that the position is not clear cut, puts forward a number of reasons in support of the proposition that the charge is a floating charge. I do not use the word “professedly” as a criticism. The fact that the OR has argued for a floating charge means I have had the benefit of adversarial argument of considerable subtlety. I agree with counsel that the issue is not straightforward and preface what follows by confessing that I have not found it easy to resolve.

The application is supported by a single witness statement of Mr Allen. Harbert has not adduced written evidence, and cross-examination has been dispensed with. The application proceeds on the basis of a list of agreed facts which appears as an appendix to this judgment.

In those circumstances I do not propose to deal with the background in detail. It is, however, helpful to set out in brief, by reference to the agreed facts, but also to other material before me, what the Company did and how it did it, as well as certain other matters, by way of backdrop to what follows.

The Company is a wholly owned subsidiary of Virtual Infrastructure Group Ltd. It provided cloud computing services, mainly to central government departments and local authorities. The provision of those services involved the use of IP addresses.

IP addresses are dealt with in paragraphs 1-7 of the list of agreed facts, and both Mr Tucker and Ms Allsop helpfully explain them in their skeleton arguments. Ms Allsop likens an IP address to the postal address of a property. It allows data to travel through a network and arrive at its intended destination. Each internet connected device has an IP address which is allocated either on a dynamic basis (the IP address changes each time the device is connected to the internet) or a static basis (the IP address is allocated for a fixed period and remains the same each time the device is connected to the internet). Mr Tucker cites passages from an article by Hilco Streambank, specialists in the disposal of such rights:

“Internet Protocol (IP) is a set of rules for addressing and routing data so it can travel thro[ugh] networks and arrive at its intended destination. Internet-connected devices each have a unique IP address.”

[…]

“Individual IP addresses are unique identifiers most often, but not always, associated with [a] specific device. Though used one at a time, they are transferred in ‘blocks’ that may inclu[de] many IP addresses. Possession of a block includes unique registrations of addresses in registries worldwide. These registries maintain uniqueness in the sense that they make [sure] it's clear who is the authorized user of a block of IP addresses. Registration includes the ability to transfer IP address use to someone else, subject to the policy of the registries, [in] return for payment.”

Internet Protocol Version 4 (IP v.4) addresses were the first major version of IP addresses and remain the dominant version. They are finite in number: there are not enough to cater for all internet connected devices; and although a successor version, IP v.6, has been developed, it is regarded as inferior to version 4. This means that IP v.4 addresses are much in demand. Millions are bought and sold every year.

In Europe, IP addresses are allocated by the Réseaux IP Européens Network Coordination Centre, one of five Regional Internet Registries which provide internet resource allocations, registration services and co-ordination in support of the operation of the internet globally.

RIPE NCC maintains a database listing the registered holders of allocated IP addresses. Only RIPE NCC members can hold IP addresses. The registered holder is known as a “responsible organisation.” The Company was, at all relevant times, a RIPE NCC member and a responsible organisation.

The terms on which a responsible organisation holds IP addresses are set out in the RIPE NCC Standard Services Agreement. I was taken to some passages in that agreement, notably the three sentence recital; clause 5, which provides for the member (in this case the Company) to make certain payments for RIPE NCC’s services; clause 9, dealing with the term of the agreement and termination; and clause 10. Mr Tucker attached some importance to clause 10.1, which restricts the member/company’s ability to assign any rights or obligations without RIPE NCC’s written consent; clause 10.2, confirming that registration “does not confer upon the Member or the third party any rights of ownership,” and the ability of RIPE NCC in certain circumstances to de- and re-register “Internet Number Resources,” which term includes IP addresses.

The importance of all this, as I understand it, is that the Company had the use of the IP addresses on very limited terms: the right to use and the right (in certain circumstances) to seek a transfer. I was referred in that respect to clause 2.0 of a document called RIPE Resource Transfer Policies which deals with transfers within the relevant service region and which again restricts the ability to transfer complete or partial blocks of addresses. (There are two iterations in evidence covering different periods, but I am told they contain the same terms as far as they are relevant for present purposes).

I should mention another document, the Company’s internal Guidance – IP Address Usage and Allocation (it is in the previous name of the Company, Skyscape Cloud Services Ltd) which deals with what is sometimes referred to as the sub-allocation of IP addresses, the Company’s right to “reclaim a customer assignment where there are a large number of IP addresses that are not being used,” and the right, where that is the case, to request a smaller allocation. Harbert, as I understand it, would not have seen this.

The Company held a number of IP addresses, the most valuable being 23,552 IP v.4 addresses. It also held a number of IP v.6 addresses, although they are not material for present purposes. It purchased IP addresses in range 83.151.208.0/20 on or around 12 May 2014 from Mailbox Internet Ltd and IP address range 51.179.192.0/18 on or around 20 May 2015 from the Department for Work and Pensions. These are the two largest ranges and represent around 87% of the total pool of IP addresses.

The Company paid an annual fee to RIPE NCC for the use of the IP addresses. Customers paid the Company for the services it provided.

The Company necessarily contracted with its customers for the provision of its services, which included, indeed necessitated, the use of one or more IP addresses. It appears from the Guidance (and has been confirmed to the special managers by a former employee of the Company) that a customer would be informed by the Company of the fact of the allocation of specific IP address(es) which would be exclusively used by the customer for the duration of the service provided. There was no written agreement dealing with the relationship between the Company and its client or customer, and the Company’s contractual documentation with its customers did not refer to the IP addresses. Although addresses were sub-allocated to customers in accordance with Company policy, the customers had no formal rights to or in respect of them: they were not, for example, granted a formal lease, licence or assignment; nor did they acquire anything in the nature of a proprietary right (see paragraph 9 above).

The Company, through its parent, raised finance governed by a loan agreement dated 28 February 2020 and made between Virtual Infrastructure and Harbert. The debenture that is the subject matter of this application was entered into on the same date.

The company was wound up by the court on 25 October 2022, the OR was appointed as liquidator, and special managers were appointed.

Thereafter the Company’s services were supported by the government whilst the OR pursued a sale and, in the alternative, a controlled migration of its customers, ultimately unsuccessfully, before terminating customer services.

The Company’s assets, including the IP addresses, have been realised. Harbert agreed to their sale subject to such security rights as it had attaching to the proceeds of sale. Harbert is owed in excess of £6 million. Whatever the outcome of this application, it will not be paid in full. The deficiency as regards unsecured creditors is estimated to be in the region of £34 million, and the value of the remaining assets in issue, the IP addresses, is estimated to be in the region of £700,000. If they are not subject to a fixed charge, their proceeds will be used to defray in part expenses incurred in the liquidation.

The label used in a debenture to describe the nature of the charge created over an asset or class of assets is a guide to the security the parties intended to create but it is not conclusive: however the parties may have chosen to describe a charge, if it is properly speaking a floating charge the court will treat it as such (see In re Spectrum Plus Ltd [2005] UKHL 41, [2005] 2 AC 680). As Edwin Johnson J said in Re Avanti Communications Ltd [2023] EWHC 940 (Ch), “The court is fundamentally concerned with the nature of the rights and obligations the parties intended to create.”

A debenture is a contract. Ms Allsop submits that in construing a contract, the court’s task is to ascertain the objective meaning of the contract, read as a whole, in the light of the background knowledge reasonably available to the parties at the time the contract was concluded. The court will consider the meaning of contractual terms by reference to (i) the natural and ordinary meaning of the relevant clause(s), (ii) any other relevant provisions of the contract, (iii) the overall purpose of the relevant clause(s) and the contract, (iv) the facts and circumstances known or assumed by the parties at the time the contract was entered into, and (v) commercial common sense, but (vi) disregarding subjective evidence of any party's intentions. The less clear the centrally relevant words to be interpreted, or the worse their drafting, the more ready the court will be to depart from their natural meaning.

Mr Tucker relies on the summary of principles given by Popplewell J in Lukoil Asia Pacific Pte Ltd v Ocean Tankers (Pte) Ltd (The Ocean Neptune) [2018] EWHC 163 (Comm), [2018] 2 All ER (Comm):

“[8] There is an abundance of recent high authority on the principles applicable to the construction of commercial documents, including Investors’ Compensation Scheme Ltd v West Bromwich Building Society, Investors’ Compensation Scheme Ltd v Hopkin & Sons (a firm), Alford v West Bromwich Building Society, Armitage v West Bromwich Building Society [1998] 1 All ER 98, [1998] 1 WLR 896; Chartbrook Ltd v Persimmon Homes Ltd (Chartbrook Ltd and anor, Pt 20 defendants) [2009] UKHL 38, [2009] 4 All ER 677, [2009] 1 AC 1101; Re Sigma Finance Corpn [2009] UKSC 2, [2010] 1 All ER 571; Rainy Sky SA v Kookmin Bank [2011] UKSC 50, [2012] 1 All ER (Comm) 1, [2011] 1 WLR 2900; Arnold v Britton [2015] UKSC 36, [2016] 1 All ER 1, [2015] AC 1619; and Wood v Capita Insurance Services Ltd [2017] UKSC 24, [2018] 1 All ER (Comm) 51, [2017] AC 1173. The court’s task is to ascertain the objective meaning of the language which the parties have chosen in which to express their agreement. The court must consider the language used and ascertain what a reasonable person, that is a person who has all the background knowledge which would reasonably have been available to the parties in the situation in which they were at the time of the contract, would have understood the parties to have meant. The court must consider the contract as a whole and, depending on the nature, formality and quality of drafting of the contract, give more or less weight to elements of the wider context in reaching its view as to the objective meaning of the language used. If there are two possible constructions, the court is entitled to prefer the construction which is consistent with business common sense and to reject the other. Interpretation is a unitary exercise; in striking a balance between the indications given by the language and the implications of the competing constructions, the court must consider the quality of drafting of the clause and it must also be alive to the possibility that one side may have agreed to something which with hindsight did not serve his interest; similarly, the court must not lose sight of the possibility that a provision may be a negotiated compromise or that the negotiators were not able to agree more precise terms. This unitary exercise involves an iterative process by which each suggested interpretation is checked against the provisions of the contract and its commercial consequences are investigated. It does not matter whether the more detailed analysis commences with the factual background and the implications of rival constructions or a close examination of the relevant language in the contract, so long as the court balances the indications given by each.”

When the issue is whether a debenture creates a fixed or floating charge, “The question is not merely one of construction” (per Lord Millett in Agnew v Commissioners of Inland Revenue [2001] UKPC 28 [2001] 2 AC 710 at paragraph 32). He continued:

A fixed charge is often easy to identify, for example when it applies to real property. In some cases it is harder to ascertain. It is useful, then, to begin by looking to some of the guidance that has been given by the courts as to what makes a charge fixed or floating.

What constitutes a fixed charge and the key distinctions between the two kinds of charge were identified by Lord Walker in Re Spectrum:

In In re Yorkshire Woolcombers Association Ltd [1903] 2 Ch 284 at page 295 Romer LJ identified a floating charge by reference to three characteristics:

In Agnew v HMRC, paragraph 13, Lord Millett said, “This was offered as a description and not a definition.” He went on:

Clause 3.2(d) of the debenture provides:

“3.2

As a continuing security for the payment and discharge of the Secured Liabilities, the Company with full title guarantee charges to the Lender by way of first fixed charge:

[…]

(d)

all licences, consents and authorisations (statutory or otherwise) held or required in connection with the Company’s business or the use of any Secured Asset, and all rights in connection with them.”

I begin with the meaning of clause 3.2(d) and its extended definition in clause 1.2(o).

Mr Tucker makes the obvious but cogent point that the clauses make no express reference to IP addresses. He is right, as he is in pointing out that the parties could easily have made express reference to them if they had intended to create a fixed charge over them. (He referred to this as the narrow interpretation.) The fact that they did not is, I accept, a pointer to the intention of the parties but not conclusive.

Ms Allsop advocates a wider interpretation which rests on the natural and ordinary meaning of the words used, contending that the IP addresses are plainly caught. This was her primary point in her oral submissions. In paragraphs 35 of her skeleton argument she also says:

It appears to me that some of the nouns used in the charging clause (“licences,” “consents,” “authorisations”) are apt to refer to giving permission, including permission to use IP addresses. Those words are, in my view, indicative of an intention to catch just the kind of assets and rights that are contemplated as arising in connection with the business carried on by the Company. The words are being used in the sense of “permission” of one kind or another; their multiplicity evinces an intention to catch assets that are, by their nature, less usual, less tangible, than those to which a debenture is traditionally directed, although that is changing fast. A licence can be a formal, legal document; but the word can also be used in a wider, less technical sense to mean little or no more than “permission.” “Consent” requires no clarification. “Authorisation,” which is the word Ms Allsop emphasised and ultimately relied on in her oral submissions, with its overtone of officialdom, also implies permission and the use of which, I agree, is appropriate to a situation where permission is given by allocation on the part of an organisation such as RIPE NCC, following which the benefit is passed on by sub-allocation, in this case by the Company to its customers.

Permission to use is also, in my view, arguably caught by the expressions “held or required in connection with the Company’s business or the use of any Secured Asset” and “all rights in connection with them,” although I accept that that is something of a stretch.

In Re Avanti, considered in greater detail below, permissions and licenses were treated as rights capable of being subject to fixed charge security, but I am cautious of giving much weight to that outcome since it was not the main issue between the parties in that case, and, of course, the debenture was not the one in this case. It gives support, however, to the ability to create a fixed charge over assets of the kind in issue here.

I have so far said nothing about “registration,” a word that is used in clause 1.2(o). It presents particular difficulties. Mr Tucker says that it is not clear how “registration” (being, he says, what is relied on as what he calls “anchor property”) could itself be subject to a fixed charge. We have seen that RIPE NCC’s Standard Services Agreement conferred no ownership rights on the Company, that the debenture makes no express reference to there being a charge over the IP addresses themselves, nor on the right to allocate them to devices, nor yet on the right to seek an assignment. Mr Tucker says there is no “anchor property” bringing with it “all other rights in connection with it” – including the right to ask for an assignment, the right that has value.

I see force in that, so that it is hard to see how registration of the kind with which we are concerned here could be subject to a charge. I agree that the inclusion of “registration” can, on one view, give rise to problems, in particular in the absence of a property right. Indeed the fact that the Company acquired no ownership or property rights gives rise to a more general difficulty affecting all the words in the debenture that might be relied on in support of the contention that they are capable of giving rise to a fixed charge. This has practical as well as drafting consequences.

That said, in my judgment, the natural and ordinary meaning of the language used in the debenture (in particular the word “authorisations”) does evince an intention to create a fixed charge on the IP addresses. But that is not the end of the matter.

Mr Tucker relies on the proposition that a clause of the kind with which we are concerned here falls to be construed as a whole, on what has been called an “all or nothing” basis: that is to say, all assets that fall within the clause must be subject to either a fixed charge or a floating charge. In support of that he relies on four cases.

In Re G E Tunbridge Ltd [1995] 1 BCLC 34 the applicant sought a declaration that 337 auction lots were subject to a fixed charge. The application failed, Sir Mervyn Davies holding the charge to be a floating charge. The judge rejected the contention that a multiplicity of different assets could be caught by a fixed charge on the basis that they fell within the same clause:

Mr Tucker’s submission gains strength if one looks at Park J’s decision in Re ASRS, which adopts the reasoning of Re G E Tunbridge. Park J said:

In Smith (Administrator of Cosslett (Contractors) Ltd) v Bridgend County Borough Council [2001] UKHL 58, [2002] 1 AC 336 it was argued that a single clause (condition 63) could be construed in two different ways to cover different kinds of asset. That argument was rejected by Lord Hoffmann:

The fourth case on which Mr Tucker relies is Re Beam Tube Products Ltd [2006] EWHC 486 (Ch) at paragraph 33 of which Blackburne J summarised the position as follows.

Ms Allsop did not suggest that the clause in issue here was not susceptible of an “all or nothing” analysis, nor did she seek to say that it was not an established approach to the construction of clauses of this kind. Her submissions came in two parts.

First, she relied on a passage in Lightman and Moss on fixed and floating charges. In para 3-013 of The Law of Administrators and Receivers of Companies (6^th edn) the authors say:

Secondly, she sought to argue, with some ingenuity, that the “all or nothing” approach in the cases cited by Mr Tucker was not part of the ratio of the decision or was otherwise not the basis on which the case was ultimately decided.

Re-reading Park J’s judgment in Re ASRS and paragraph 33 of Blackburne J’s judgment in Re Beam Tubes, I have to say that I am not convinced: Blackburne J seems firm: “In my view it is all or nothing” reads to me as unequivocal and as being a general proposition. Park J similarly expressed himself in the clearest terms:

Having dealt with the language of the debenture, I move to the submissions on the nature of IP addresses.

The first is whether they are of a nature that makes them susceptible of being subject to a fixed charge or only a floating charge.

If an asset is part of the company’s circulating capital or is a fluctuating asset or body of assets it is more likely that any charge to which it is subject will be floating rather than fixed. That was the second characteristic identified in Yorkshire Woolcombers. In Re Spectrum Lord Walker said:

Mr Tucker contends that the IP addresses formed part of the Company’s circulating capital. Ms Allsop argues the contrary.

I go some of the way with Ms Allsop on this point. Whilst there is no hard and fast definition, a feature of circulating capital is that it is disposed of and replaced. That is clear from the passage from Re Spectrum above and from other authority. The IP addresses with which we are concerned do not readily fall under the “circulating” or “fluctuating” description. But I am unable to go the whole way with Ms Allsop. In Re Spectrum Lord Walker said, “The chargee has a proprietary interest, but its interest is in a fund of circulating capital…” The Company here has and had no proprietary interest in the addresses, and its capacity to deal with them was limited: it could sub-allocate them, withdraw them in certain circumstances, reassign them and transfer them; but not simply sell or dispose of them and then replenish the gap created as would be the case with stock or some other asset dealt with as part of the “churn” of business conducted by a shop or similar concern. The IP addresses do not seem to me to be amendable of easy categorisation They seem in some respects to be like the shares in Arthur D Little Ltd v Ableco Finance LLC [2002] EWHC 701 (Ch) which were held not to be part of the chargor’s circulating capital as the chargor did not need to sell them, deal with them, or substitute them as part of its ordinary business. But that is just one aspect of the case. Although the debenture in that case was held to have created a fixed charge, I do not think it follows that the IP addresses in this case fall to be treated in the same way. It cannot be said that the Company was able to trade in the IP addresses in the ordinary way (cf Lord Walker above), and the analogy with the shares breaks down in light of the fact that in Arthur D Little the charge provided for the chargee to hold the certificates and other documents of title as well as signed, undated stock transfer forms. There was something over which they could exercise real control. There is no analogous control in this case; indeed it is hard to see how the control mechanisms relied on by Harbert could apply at all in practice to the assets in issue here. I shall come to actual control below. 

I have concentrated above on the “circulating capital” description used in the case law. “Fluctuating body of assets,” another phrase used in the case law, is arguably wider, but there was negligible fluctuation in this case, again by reason of the limits to what the Company could do, so I do not think this takes things any further on the facts of this case.

The fact that the IP addresses cannot have been part of the Company’s circulating capital and did not fluctuate in the sense used in the authorities means that one characteristic commonly associated with the creation of a floating charge is absent in this case; but it does not follow from that that the charge applicable to them was fixed. I regard the point as inconclusive in this case.

The same cannot be said of control. This is a convenient point at which to look at the issue of control and the decision in Re Avanti in more detail.

Four classes of assets were considered by Edwin Johnson J in Re Avanti, but the one which invites particular attention for present purposes took the form of satellite network filings registered with the International Telecommunications Union. The filings were explained by the judge in paragraph 9 of his judgment. In summary, in accordance with international treaties, the ITU maintains a database of all registered satellite sand terrestrial frequency assignments (the Master International Frequency Register). Ofcom, on behalf of the UK, decides on applications for a satellite network filing. If an application is approved and registered on the MFIR, the registrant acquires the right to use the registered filing. A filing recorded as held by a UK satellite operator may, subject to conditions, be transferred with the consent of Ofcom to another UK operator. The Avanti group of companies, of which Avanti Communications Ltd was one, operated satellites within that framework. The company (and its holding company) went into administration. It sold substantially all its business, including the relevant assets, on the basis that they were subject to a fixed charge.

Clause 3 of the debenture was the clause by which the company charged various categories of assets which the debenture went on to describe. Clause 3.1(b)(ix) provided a charge over “the benefit of all licences, consents and agreements held by it in connection with the use of any of its assets.” “This definition,” the judge said, “seems to me to have been wide enough to include the Satellite Network Filings […]. I did not understand it to be in dispute that the Satellite Network Filings and the PES Licenses, as permissions and licenses, were rights capable of being subject to fixed charge security.”

The judge disposed of the first stage of the test to be applied by reference to two questions:

As we have seen, in the instant case the debenture imposed restrictions but also permitted disposals.

Edwin Johnson J cited Lightman and Moss (paragraph 3-021):

He went on, at paragraph 118, to accept that it was not the case that only a total prohibition on dealing with an asset made it possible for it to be subject to a fixed charge, rejecting the views of textbook writers to that effect. He said:

Millett LJ (with whom Lord Hoffman on appeal agreed) in Re Cosslett put it like this:

That leads to the question of control in this case.

Mr Tucker says that his client struggles with the notion that Harbert had sufficient control. He relies on the following (paragraph 73 of his skeleton argument):

Ms Allsop submits that post-contractual conduct is generally irrelevant and inadmissible in this context (per Edwin Johnson J in Re Avanti paragraph 38), although the nature of the assets in question may also be taken into account (ibid paragraph 35), and regard may be had to the nature of the business of the charger (ibid paragraph 37).

I respectfully follow the “nuanced” approach advocated by Edwin Johnson J, holding the exercise of control to be one of the factors I am entitled to take into account. He said:

As to the exercise of control in this case, I note again that the terms of the debenture provide for control to be exercised. I find some difficulty, however, in concluding, as a matter of fact, that it was exercised. Mr Allen asserts (or implies) in his witness statement that it was not, laying down the gauntlet and inviting a challenge.

Contractual control obligations cannot be ignored. The Company was bound by its obligations to Harbert until or unless they were relinquished or waived. There is no evidence of either having occurred, nor was that submitted against Harbert, so the covenants to which the Company was subject obtained throughout its relationship with Harbert. If the Company failed to comply with its obligations, on one view that is to be laid at its door; it is not a default on the part of the chargee, which in any event is not obliged to enforce its strict legal rights at any particular time as a condition of preserving them. Failure to do so may, however, be indicative of how the parties viewed the charge, as Mr Tucker contends.

He also submits that the effect of the charging clause relied on would have been, potentially, to bite on a number of assets that cannot realistically have been in the parties’ contemplation. As he puts it, the Company would necessarily have approved, authorised, consented to or resolved to enter into all kinds of contracts, licences and so on, its rights to which would, on the wide interpretation contended for by Harbert, be subject to a fixed charge. For example, the Company must have used software in the course of its business which must have been updated and replaced from time to time, often alongside the hardware on which it has been installed. Software is usually supplied under the terms of a user licence. Mr Allen produces a list of nine such items. Mr Tucker submits that it cannot sensibly be said that each version of Windows, and any and all other commercial software licenced to the Company, or even the frequently changed laptops on which it was installed, was susceptible of control and the subject of a fixed charge.

The point is a nice one, if only to the extent that it indicates a failure of the parties to consider the true effect of what they were supposedly agreeing; but it is not as powerful as the absence of evidence of control over the assets in issue and others of the kind canvassed in the previous paragraph.

In my view, in the absence of any challenge to Mr Allen’s evidence I am bound to conclude that Harbert did not exercise control or seek to do so. Sometimes the absence of evidence is as powerful as the presence of evidence. This is such a case. It is also a case where some regard should be had to post-contractual conduct. The control provisions in the debenture were, then, a “sham,” not in any fraudulent sense, but in the sense used in Re Avanti. In reaching that view I also take into account the apparent ability of the Company to carry on its business without the consent of Harbert (see the passages from Agnew v HMRC, Re ASRS and Re Spectrum cited in paragraphs 25, 27 and 28 above).

On that basis and for the other reasons given above, in particular those on the “all or nothing” principle, I conclude, albeit not without misgivings, that the charge in this case is a floating charge. I reach that conclusion by reference to both stages of the process identified in Agnew v HMRC.

I am grateful to Mr Tucker and Ms Allsop for skeleton arguments and oral argument of the highest quality and for their patience in the course of the latter with my frequent interruptions and questions as I tried to come to grips with law and submissions that I have found to be far from easy. I am also grateful to Mr Allen and his advisers for his helpful and carefully considered written evidence and to the applicant’s solicitors for an immaculately prepared hearing bundle.

I will hear counsel on the form of order I should make when this judgment is handed down.