First-tier Tribunal (Tax)

[2018] UKFTT 27 (TC)

[image removed]

TC06306

Appeal number:   TC/2012/04698

 

VAT  - recovery of input tax on suppliers of broking, claims handling and underwriting support services  - appeal allowed

 

 

FIRST-TIER TRIBUNAL

TAX CHAMBER

 

 

 

	HASTINGS INSURANCE SERVICES LIMITED	Appellant
	- and -
	THE COMMISSIONERS FOR HER MAJESTY’S	Respondents
	REVENUE & CUSTOMS

 

 

 

TRIBUNAL:	JUDGE HARRIET MORGAN MEMBER  MR NIGEL COLLARD

 

 

Sitting in public at the Royal Courts of Justice, the Strand, London on 14 to 18, 21, 23 and 24 November 2016

 

 

Mr Andrew Hitchmough QC and Ms Barbara Belgrano, counsel, instructed by Ashurt LLP, for the Appellant

 

Mr Raymond Hill, instructed by the General Counsel and Solicitor to HM Revenue and Customs, for the Respondents (“HMRC”)

 

 

 

 

 

© CROWN COPYRIGHT 2018

DECISION

 

 

1.              The appellant is an insurance services company operating in the UK.  The appeal relates to whether, for VAT purposes, the appellant can recover or obtain credit for (“recover”) input tax it incurred in the period from 1 February 2009 to 31 December 2013 (the “Period”), which is attributable to supplies of broking, underwriting support and claims handling services (the “services”) made to Advantage Insurance Company Limited (“Advantage”). 

2.              Advantage is a company related to the appellant based in Gibraltar which underwrites UK private and commercial motorcar and motorcycle insurance.  During the Period Advantage made supplies of insurance to UK customers, acting through the appellant as its broker or intermediary, and using the other services provided by the appellant.

Overview

3.              The applicable law is set out at [13] to [29].  In summary, it was common ground that, if the supplies of the services were made in the UK, they fall within the exemption from VAT for certain insurance services (under item 4 of group 2 of schedule 9 to the Value Added Tax Act 1994 (“VATA”)).  A taxable person, such as the appellant, is not entitled to recovery of input tax attributable to exempt supplies made in the UK.  However, the appellant is entitled to recovery of the input tax if the supplies of services were in fact made to a taxable person which “belongs” outside the EU member states.  Advantage was at all relevant times a taxable person.  Gibraltar was not and is not an EU member state for VAT purposes. The dispute was whether Advantage “belongs” outside the EU or not for this purpose.

4.              Where a person “belongs” for this purpose is set out in the UK place of supply rules.  These rules changed during the Period but, in effect, under both sets of the applicable rules, a taxable person “belongs” for this purpose:

(1) where it has a business establishment (“BE”); or

(2) if different, where it has a fixed establishment (“FE”); or

(3) if it has both a BE and a FE (or several such establishments), where the establishment is located which is most directly concerned with the supply.

5.              It was not disputed that Advantage had a BE in Gibraltar.  The question was whether it also had a FE in the UK and, if so, whether the supplies of services were made to that FE rather than to its BE in Gibraltar.  The place where Advantage’s supplies of insurance were made was also relevant to the analysis and is also to be determined by reference to where Advantage “belongs”.  This similarly depends on whether Advantage made the supplies of insurance from its BE in Gibraltar or from a UK FE.  Those supplies also fall within the insurance exemption referred to in [3] above if made in the UK.

6.              These UK rules are to be interpreted in accordance with the relevant EU VAT law, which the UK rules are intended to enact, as interpreted by the Court of Justice of the European Union (“CJEU”).  The CJEU has considered in a number of cases the question of when, for the purposes of deciding in which country a supply is made (and so which country has taxing rights), a FE may be held to exist and when it is appropriate to allocate taxing rights to any such FE rather than to a taxable person’s BE.  The CJEU has consistently held that the place where the BE is located is the primary point of reference unless that gives an irrational result.  The case law is set out at [407] to [482].

7.              There was no definition or further explanation of the terms BE and FE in the UK or the EU rules until the introduction in March 2011 of Council Implementing Regulation 2011 (282/2011/EU) (the “Regulation”). The Regulation includes a provision defining a FE, for the purposes of the place of supply rules, as based on the previous case law in the CJEU, as an establishment other than a BE:

“characterised by a sufficient degree of permanence and a suitable structure in terms of human and technical resources”, where looking at the location of the recipient of the supply, “to enable it to receive and use the services supplied to it for its own needs” or, where looking at the location of the supplier, “to enable it to provide the services which it supplies”. 

8.              The parties took different views on the correct application of the FE test in the Regulation and the interpretation of the FE test set out in the cases.  (The case law remains relevant, as the terms of the Regulation which reflect the test as set out in the cases suggest and as has been confirmed in the CJEU more recently.) 

9.              In summary, the appellant’s stance was that the input tax attributable to the services is recoverable as the services were supplied to Advantage at its BE in Gibraltar.  This was on the basis that Advantage made its supplies of insurance to UK customers from the BE in Gibraltar (acting through the appellant as broker) and, accordingly, that it received the services to enable it to do so at that BE.

10.           HMRC considered that the input tax was not recoverable on a two stage analysis:

(1) First, they submitted that Advantage’s supplies of insurance services to UK customers were made in the UK through a FE comprising the appellant’s human and technical resources.  In their view it sufficed for there to be a FE that, under the contractual arrangements, the marketing and sale of insurance and related “customer facing” activities were carried out by the appellant, on behalf of Advantage, through the appellant’s staff and systems in the UK, in circumstances where the conduct of the appellant’s business was subject to “approvals” from Advantage. 

(2)  Secondly, it follows from that, they argued, that the appellant’s supplies of services were made to Advantage at that FE in the UK, as the establishment at which the services supplied by the appellant were received and used by the FE for its own needs in order to enable it to make the insurance supplies. Whilst primacy is generally to be accorded to the BE, in HMRC’s view, this was a case where to do so would give an irrational result potentially leading to distortion of competition. 

11.           The appellant countered that Advantage simply did not have a FE in the UK through which it supplied insurance or at which it could receive and use the services for its own needs.  In its view, for the appellant’s resources to comprise a FE:

(1) Advantage would need to have control over the resources comparable to that of an owner.  In its view, Advantage plainly did not have that level of control.

(2) The resources would need to provide in the UK all that was necessary for Advantage to make the insurance supplies or receive the services from the appellant for its own needs and use (as relevant).  The appellant noted that in fact many of the functions needed for Advantage to make insurance supplies were carried out through its own staff and resources in Gibraltar, such as making the underwriting decisions (in particular, deciding what to insure for what risk price) and decisions on large loss claims and dealing with reinsurance, investment and regulation.

12.           The appellant considered that, if that is not correct and there was a FE in the UK, in any event, only the “customer facing” services were received and used at any such FE and the remainder were received and used at Advantage’s BE in Gibraltar.  In the appellant’s view, in such circumstances, there is no reason to depart from the BE as the place of supply/belonging.

Law – position with effect from 1 January 2010

UK input tax rules

13.           Under the rules applicable from 1 January 2010, a taxable person has the right to credit for input tax attributable to supplies made by him in the course of furtherance of his business which is attributable to (a) taxable supplies (b) supplies outside the UK which would be taxable supplies if made in the UK and (c) such other supplies outside the UK and such exempt supplies as the Treasury may by order specify for this purpose (under s 26 (1) and (2) VATA).  The supplies of services in this case fall within a special case designated under (c) as set out below.

14.           The following are specified to fall within (c) above under article 3 of the Value Added Tax (Input Tax) (Specified Services) Order 1999:

“3. Services–

(a) which are supplied to a person who belongs outside the Member States;

(b) which are directly linked to the export of goods to a place outside the member States; or

(c) which consist of the provision of intermediary services within the meaning of item 4 of Group 2, or item 5 of Group 5, of Schedule 9 to the Value Added Tax Act 1994 in relation to any transaction specified in paragraph (a) or (b) above,

provided the supply is exempt, or would have been exempt if made in the United Kingdom, by virtue of any item of Group 2, or any of items 1 to 6 and item 8 of Group 5, of Schedule 9 to the Value Added Tax Act 1994.”

15.           These rules implement article 169 of Council Directive 2006/112/EC (the “Directive”) which provides that a taxable person can deduct VAT on supplies which would otherwise be exempt under article 135(1), which includes supplies performed by insurance agents, where the customer is established outside the EU.

UK place of supply rules

16.           The UK place of supply rules set out where a person “belongs” for the purposes of establishing in which country a supply is made and which country has taxing rights.  Under these rules, the place where a supply of services is regarded as taking place is:

(1) where “the person to whom the services are supplied is a relevant business person, in the country in which the recipient belongs”, and

(2) “otherwise, in the country in which the supplier belongs” (under s 7A(2) VATA). 

17.           In this case, it is not disputed that supplies of insurance made by Advantage to UK customers were made where it belonged as the supplier.  As it is accepted that Advantage is a relevant business person, the supplies of the services made by the appellant to Advantage were made where Advantage belonged as the recipient of the services.

18.           A relevant business person is treated as belonging in the country where it has a BE or some other FE, where it has none in any other country, or, if the person has a BE or some other FE or FEs in more than one country, in the country in which the establishment is located which is “most directly concerned with the supply” (s 9(2) and (3) VATA).  A person who is not a relevant business person is to be treated as belonging in the country in which the person’s usual place of residence is (s 9(5) VATA).

EU place of supply rules

19.           The above UK rules are intended to implement articles 44 and 45 of the Directive which provide, so far as is material, that:

[44] “The place of supply of services to a taxable person acting as such shall be the place where that person has established his business. However, if those services are provided to a [FE] of the taxable person located in a place other than the place where he has established his business, the place of supply of those services shall be the place where that [FE] is located…”

[45] “The place of supply of services to a non-taxable person shall be the place where the supplier has established his business.  However, if those services are provided from a [FE] of the supplier located in a place other than the place where he has established his business, the place of supply of those services shall be the place where that [FE] is located…”

Definition of terms in the Regulation

20.           As noted the terms FE and BE are defined in articles 10 and 11 the Regulation which was introduced on 15 March 2011.  The preamble to the Regulation states that it is binding and directly applicable in all member states from the date it came into force in 2011.  Its stated aim is to ensure uniform application of the current VAT system by laying down rules implementing the VAT Directive, including the place of taxable transactions (para 4).  It contains specific rules in response to selective questions of application which are designed to bring uniform treatment to those specific circumstances only and are not conclusive for other cases and, in view of their formulation, are to be applied restrictively (para 5).  The preamble also states that concepts such as BE and FE should be clarified “to ensure the uniform application of rules relating to the place of taxable transactions” and, “whilst taking into account the case law of the CJEU, the use of criteria which are as clear and objective as possible should facilitate the practical application of these concepts”.

21.           Article 10 of the Regulation provides that:

“1. For the application of Articles 44 and 45 of [the Directive], the place where the business of a taxable person is established shall be the place where the functions of the business’s central administration are carried out.

2. In order to determine the place referred to in paragraph 1, account shall be taken of the place where essential decisions concerning the general management of the business are taken, the place where the registered office of the business is located and the place where management meets.…” [As noted already, in the case of Advantage it is common ground that this place is Gibraltar.]

22.           Article 11 of the Regulation (“article 11”) provides that:

“1. For the application of Article 44 of [the Directive], a 'fixed establishment' shall be any establishment, other than the place of establishment of a business referred to in Article 10 of this Regulation, characterised by a sufficient degree of permanence and a suitable structure in terms of human and technical resources to enable it to receive and use the services supplied to it for its own needs.

2.  For the application of the following Articles [including Article 45], a 'fixed establishment' shall be any establishment, other than the place of establishment of a business referred to in Article 10 of this Regulation, characterised by a sufficient degree of permanence and a suitable structure in terms of human and technical resources to enable it to provide the services which it supplies.” (emphasis added)

23.           The Regulation also addresses the question of how a supplier should determine the place of supply where an entity has a BE in one country and a FE (or FEs) in one or more other countries. 

(1) Article 20 deals with the case where there is a single establishment in one country only, such that the supply of services is taxable only in that country. 

(2) Under article 21:

“Where a supply of services to a taxable person…falls within the scope of Article 44 of [the Directive], and the taxable person is established in more than one country, that supply shall be taxable in the country where that taxable person has established his business.

However, where the service is provided to a [FE] of the taxable person located in a place other than that where the customer has established his business, that supply shall be taxable at the place of the [FE] receiving that service and using it for its own needs. …”

(3) Under article 22:

“1. In order to identify the customer’s [FE] to which the service is provided, the supplier shall examine the nature and use of the service provided.

Where the nature and use of the service provided do not enable him to identify the [FE] to which the service is provided, the supplier, in identifying that [FE], shall pay particular attention to whether the contract, the order form and the VAT identification number attributed by the Member State of the customer and communicated to him by the customer identify the [FE] as the customer of the service and whether the [FE] is the entity paying for the service.

Where the customer’s [FE] to which the service is provided cannot be determined in accordance with the first and second subparagraphs of this paragraph or where services covered by Article 44 of [the Directive] are supplied to a taxable person under a contract covering one or more services used in an unidentifiable and non-quantifiable manner, the supplier may legitimately consider that the services have been supplied at the place where the customer has established his business....”

Law - position prior to 1 January 2010

Input tax recovery

24.           The UK rules for obtaining credit for or recovery of input tax were the same as those set out above. 

UK place of supply rules

25.           The UK general place of supply rule was that a supply was made (a) in the UK if the supplier belonged in the UK or (b) in another country (and not in the UK) if the supplier belonged in that other country (s 7(10) VATA).  That was subject to any order or rule made by the Treasury varying this general rule (s 7(11)). 

26.           Supplies of insurance made by Advantage fell within the general rule that they were made in the UK if Advantage belonged there and otherwise in another country if it belonged there.  The supplies of services by the appellant to Advantage were made where Advantage belonged under a variation to the general rule.  This provided that supplies of this kind were treated as made where the recipient of the supply belonged if (a) the recipient belonged outside the member states (or in a country other than that in which the supplier belonged) and (b) the person received the supply for business purposes and was not treated as having supplied the services himself (under paras 13 and 16 of the Value Added Tax (Place of Supply of Services) Order 1992 and para 5, schedule 5 VATA).

27.           The UK rules determining the place of belonging were framed in similar terms to those applicable from 1 January 2010:

(1) The supplier of services was treated as belonging in a country if -

(a)    he had there a BE or some other FE and no such establishment elsewhere; or

(b)   he had no such establishment (there or elsewhere) but his usual place of residence was there; or

(c)    he had such establishments both in that country and elsewhere and the establishment of his which was most directly concerned with the supply was there (s 9(2)(a) to (c) VATA).

(2) Similarly, the recipient of a supply (other than an individual who received the supply otherwise than for business purposes) was treated as belonging in a country if (a) he had a BE or FE there and no such establishment elsewhere or he had no such establishment there but his usual place of residence was there or (b) he had such establishments as are mentioned in s 9(2) in that country and elsewhere and the establishment of his at which, or for the purposes of which, the services were most directly used or to be used was in that country (s 9(4) VATA).

28.           For the purposes of this section it was provided that “a person carrying on a business though a branch or agency in any country shall be treated as having a business establishment there” (s 9(5) VATA).   

EU VAT rules

29.           The UK place of supply rules set out above implemented the provisions in articles 43 and 56(e) of the Directive as in place before 1 January 2010 as follows:

(1) Under article 43:

“The place of supply of services shall be deemed to be the place where the supplier has established his business or has a [FE] from which the service is supplied, or in the absence of such a place of business or [FE], the place where he has his permanent address or usually resides.”

(2) By way of exception to the above general rule, under article 56(e):

“The place of supply of the following services to customers established outside the Community, or to taxable persons established in the Community but not in the same country as the supplier, shall be the place where the customer has established his business or has a [FE] for which the service is supplied, or, in the absence of such a place, the place where he has his permanent address or usually resides: … (e) banking, financial and insurance transactions, including reinsurance, with the exception of the hire of safes.”

 

 

Evidence

30.           We have found the facts on the basis of the evidence given by the witnesses of whom details are out below and the bundle of documents produced to the tribunal.

Witnesses of fact

31.           We received evidence from the following witnesses who attended the hearing and were cross-examined.  We found them all to be credible witnesses and have accepted their evidence:

(1) Mr Keith Charlton, who was a director and Chief Executive Officer (“CEO”) of Advantage from 2007 to January 2014.  Mr Charlton’s present position at Advantage is non-executive director. 

(2)  Mr Ian Godfrey, who was employed at the appellant as an underwriter from 1996 and underwriting director from 1999.  He was a director of the appellant from 2002 until January 2008 and of Advantage from 2002 until 2 February 2009 (during that period he had no other role at Advantage).  On 1 April 2013 he became the managing director of Advantage and effectively took over Mr Charlton’s role when he retired as CEO in January 2014.

(3) Mr Gary Eagar who had been Advantage’s claims director from 3 February 2009.  The team is run on a day to day basis by Ms Michelle McKeever, the head of large loss and claims development, who reports to Mr Eagar. 

(4) Mr Pavey who from February 2009 was the finance director of the appellant and a director. 

(5) Ms Lucy Johnson who joined the appellant in February 2013 as the commercial director employed on the retail and broking side of the appellant’s business. 

(6) Mr Michael Lee who has been the managing director of insurer services at the appellant since June 2011. 

(7) Mr Tom Gumbrell who joined the appellant in 1999 and during the Period was an account management analyst. 

32.           We also received a witness statement from Mr Michael Doidge who joined the appellant in March 2014 as head of technical claims.  As he did not work for the appellant during the relevant Period we did not accord relevance or weight to his evidence except to the extent it was consistent with the evidence of the other witnesses or is supported by the documentary evidence. 

Expert reports

33.           The appellant produced an expert report from Mr Chris Johnson, a non-life insurance professional since 1979 with experience of cross border insurance operations and, in particular, of regulatory practice in Gibraltar.  His report related to the legal and regulatory framework in Gibraltar, market practice for insurance companies outsourcing arrangements in Gibraltar and comments on the compatibility of the arrangements between Advantage and the appellant with Advantage’s regulatory position in Gibraltar and in relation to customers to whom it supplied insurance.  Mr Johnson attended the hearing and was cross-examined on the report.

34.           HMRC sought to produce a report from, Mr David Kendall, a UK solicitor with experience of the law relating to the regulation of the insurance industry in the UK.  The appellant applied for his evidence to be excluded.  We decided to exclude the majority of the report for the reasons set out below.  Mr Kendall attended the hearing and was cross-examined on the sections of his report which were admitted as evidence.

35.           We note that by the end of the hearing it became clear that neither party considered the regulatory position to be relevant to the question of whether Advantage had a FE in the UK through which it made supplies of insurance services and at which it received the appellant’s services.  Much of the reports of both witnesses were, therefore, in any event irrelevant.

Application to exclude the evidence of Mr Kendall

36.           Mr Kendall submitted two reports which contained the following:   

(1) Mr Kendall stated the opinions he gave were on insurance regulation and practice.  He then listed the questions he was asked to answer as follows.

(a)    Whether Advantage supplied insurance to UK policyholders from Gibraltar or through the appellant in the Period.

(b)    Did the appellant have the human and technical resources needed to supply insurance in the retail motor market in the UK and were those resources in effect the resources of Advantage for the purposes of Advantage’s supply of insurance?

(c)    Whether the services supplied by the appellant to Advantage were supplied in the UK or Gibraltar?

(2) His conclusions, as set out in summary at the beginning of the report, were: 

(a)    Advantage supplied insurance to UK policy holders from the UK through the offices of the appellant in the UK principally at Bexhill.  Without the human and technical resources at Bexhill, Advantage would not have been able to supply insurance.  He based his conclusion on the relevant contracts and the circumstances surrounding the making and performance of those contracts set out in the witness statements and documents he had reviewed.  He had also taken account of the regulatory framework.

(b)   From a commercial perspective the most significant activities involved in the conduct of insurance business are the effecting of contracts of insurance and the carrying out of those contracts (such as receiving premium, adjusting and paying claims).  Any person carrying out those activities in the UK must be authorised to do so by the relevant regulatory authority.

(c)    Correspondingly Advantage did not effect or carry out such insurance contracts in Gibraltar.  He also commented on the timing of when he considered a contract of insurance was made. 

(d)   Given the extensive delegation of the conduct of Advantage’s insurance business to the appellant within the framework of the applicable agreements Advantage exercised a degree of control over the appellant. 

(3) He set out his understanding of a number of industry terms from a legal perspective and an outline of the regulatory regime in the UK.  He set out an analysis of Advantage’s business, its regulatory status, the relationship between the parties, the services agreements and an extensive section on the appellant’s operations. 

(4) He then set out his view on whether Advantage had a branch in the UK under the European Commission’s Interpretative Communication 2000/C43/03 (“ECIC”).  This included consideration of whether, for those purposes, the appellant was subject to the direction and control of Advantage, whether the appellant was able to commit Advantage and whether the appellant had a “permanent brief”.

(5) Mr Kendall’s second report largely addressed points made in Mr Johnson’s report.

37.           The appellant applied to exclude Mr Kendall’s evidence under rule 15 of the rules governing the tribunal (the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009 (S.I. 2009/273 (L.1)) (the “Rules”)).  This includes a provision that the tribunal may give directions on the nature of the evidence or submissions it requires and whether the parties are permitted or required to provide expert evidence.  It also provides that the tribunal may (a) admit evidence whether or not the evidence would be admissible in a civil trial in the UK or (b) exclude evidence that would otherwise be admissible in certain circumstances including where it would otherwise be unfair to admit the evidence.

38.           The application was made on the following inter-related grounds:  

(1) The reports contained a substantial amount of inadmissible legal opinion.  Such opinion is not “evidence” within the meaning of rule 15 but is properly to be addressed by way of submissions on the law, as recognised by the tribunal in Deloitte LLP v Revenue and Customs Commissioners [2016] UKFTT 479 (TC), (see [49] to [51]).  It is not disputed that is consistent with the approach the courts generally take in civil cases.  The appellant cited a number of authorities including JP Morgan Chase Bank v Springwell [2006] EWHC (Comm) (at [29] to [32]). 

(2)  In any event the matters covered do not constitute admissible “expert” evidence under the principles identified by the Supreme Court in Kennedy v Cordia (Services) LLP [2016] UKSC 6 as set out below.  Although the tribunal is not bound by the normal civil rules of evidence, it did not appear to be disputed by HMRC that these principles are relevant.  The appellant noted that in Deloitte the tribunal seemed to accept that the general approach in civil cases is relevant as the tribunal considered and applied Kennedy and JP Morgan (see [52]).

(3) It is not in the interests of justice and fairness for the evidence to be admitted.  If admitted the tribunal would have to make findings of fact on the basis of the reports.  There would be a real danger that in doing so the tribunal would transmogrify the legal conclusions in the report into findings of fact which would be very difficult to challenge on appeal given the high threshold for doing so.

(4) The test for whether Advantage had a branch in the UK for regulatory purposes by reference to the ECIC is an entirely different test from the test set out by the CJEU for what constitutes a FE, as now incorporated in regulation 11.  Evidence on that is, therefore, irrelevant.

39.           In the Kennedy case the Supreme Court noted, at [39] to [41], that, not only can skilled witnesses give evidence of their opinions, they can also give skilled evidence of fact:

“based on his or her knowledge and experience of the subject matter, drawing on the work of others, such as the findings of a published research or the pooled knowledge of a team of people with whom he or she works”. 

40.           They said that the special rules that govern admissibility of expert opinion evidence also cover such skilled evidence of fact, being, as set out at [44]:

“(i) whether the proposed skilled evidence will assist the court in its task;

(ii) whether the witness has the necessary knowledge and experience;

(iii) whether the witness is impartial in his or her presentation and assessment of the evidence; and

(iv) whether there is a reliable body of knowledge or experience to underpin the expert’s evidence.”

41.           They continued, at [46], that strict necessity cannot be the test for the admissibility of skilled evidence of fact as otherwise the court could be “deprived of the benefit of a skilled witness who collates and presents to the court in an efficient manner the knowledge of others in his or her field of expertise”.  There may be circumstances in which a court could determine a fact in issue without an expert collation of relevant facts if the parties called many factual witnesses at great expense such that a strict necessity test would not be met.  They referred to a decision of the United States Supreme Court (Daubert v Merrell Dow Pharmaceuticals Inc (1993) 509 US 579) on a rule of evidence which, in their view, is consistent with the approach of Scots law in relation to skilled evidence of fact:

“If scientific, technical or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise.”

42.           They concluded, at [47], that the advantage of the formula in this rule is that:

“it avoids an over-rigid interpretation of necessity where a skilled witness is put forward to present relevant factual evidence in an efficient manner rather than give an opinion explaining the factual evidence of others.  If skilled evidence of fact would be likely to assist the efficient determination of the case, the judge should admit it”.

43.           They continued, at [48] and [49], that an expert must explain the basis of his or her evidence “when it is not personal observation or sensation”; mere assertion “is worthless”.  In addition, expert witness evidence “cannot usurp the functions of the jury or the judge sitting as a jury”. Thus, while on occasion, in order to avoid illusive language, the skilled witness may have to express his or her views in a way that addresses the ultimate issue for the court, expert assistance does not extend to supplanting the court as the decision-maker.  The fact-finding judge cannot delegate the decision-making role to the expert.  Finally, at [51] they noted that the court may exclude expert evidence which “on its face does not comply with the recognised duties of a skilled witness to be independent and impartial”.  In their view that requirement was “one of admissibility rather than merely of the weight of the evidence….”  

44.           The appellant argued that Mr Kendall’s evidence was not skilled evidence of fact as described by the Supreme Court.  He gave an opinion explaining the factual evidence of others, in applying English law to facts set out by the witnesses of fact or in the documents.  That opinion is not admissible as it does not meet the strict necessity test.  The tribunal itself has the expertise to evaluate the evidence before it, with the benefit of cross-examination of the witnesses (which was not of course available to Mr Kendall when he wrote his report) and to construe the relevant contracts in the light of the parties’ respective submissions.  The appellant said that was particularly so as Mr Collard worked for many years in the insurance industry.

45.           The appellant noted that the “usurpation” principle referred to in Kennedy was recognised in the Deloitte case where the tribunal drew a distinction between opinions “whose foundation is built on matters which are outside the tribunal’s expertise” and opinions “whose foundation itself rests on legal matters which are properly for the tribunal to reach a conclusion on with the benefit of the relevant facts and legal submission” (see [52] and [65]).  The appellant submitted that in this case Mr Kendall gave his opinion on those very legal matters which are properly for the tribunal to reach a conclusion on.  The questions put to Mr Kendall by HMRC were couched in the language of VAT and are the very questions to be determined by the tribunal in this case.  In answering those questions Mr Kendall clearly usurped the function of the tribunal, for example, in concluding that Advantage supplied insurance to UK policy holders from the UK through the offices of the appellant. 

46.           Moreover, in so far as the ECIC criteria are relevant to the proceedings, it is for the tribunal to determine their relevance on the basis of the submissions made.  Mr Kendall’s evidence again clearly usurped the function of the tribunal in that respect.  The appellant asserted that was not the case as regards Mr Johnson’s report as he addressed industry practice in Gibraltar; evidence on practice may be of assistance (see Deloitte at [54] and [56]). 

47.           In the appellant’s view, the only area in which Mr Kendall may be said to give evidence of industry practice was in his statements as regards the significant activities involved in insurance as a commercial matter.  However, this is not an area which he was asked by HMRC to give evidence on and, as a lawyer who advises insurance businesses, he had no direct experience of running such businesses. The statements were mere assertion.  The appellant also pointed to a number of other statements which in its view were mere assertion such as that “in practice and by agreement….Advantage’s brief to Hastings was exclusive”. 

48.           Finally, the appellant submitted that Mr Kendall’s treatment of the factual evidence raised serious doubts on his impartiality and independence. The appellant also noted that Mr Kendall purported to make a number of findings of fact which was held to be impermissible in the JP Morgan case (see [21]).  It is clearly for the tribunal to make findings of fact and, in any event, a number of the findings made by Mr Kendall were incorrect.

HMRC’s submissions

49.           HMRC responded that evidence is not rendered inadmissible if it merely sets the scene with undisputed legal matters.  Rather the problem arises where the evidence seeks to address disputed legal matters which thereby may be said to usurp the judge’s function.  This was recognised in Deloitte where, at [67], the tribunal said that it was “contested matters of law” which are “more efficiently addressed through submission rather than expert medium” and noted that expert reports are “not rendered inadmissible because they refer to legislation, matters of law or indeed the very issue before the court or tribunal.”  

50.           HMRC said that in referring to the terms of insurance law at the start of his report Mr Kendall was setting the scene so that it is clear how he was using the terminology in the report.  Other than as regards the term “underwriting” none of the points he made are in dispute.  Mr Kendall then made it very clear that his opinion was based on insurance regulation and practice and not on the VAT issue in dispute.  HMRC framed their questions as they did simply to ensure that the report was focussed correctly on the issues.

51.           In HMRC’s view Mr Kendall complied with the approach set out at [21] of JP Morgan that experts should not attempt to make findings of fact but “should express their opinion on the areas in which they have their expertise, on the basis of assumed facts, which should be clearly identified and stated in their expert report.”  They asserted that in his first report Mr Kendall set out the assumed facts which he believed were relevant to the regulatory issue he was dealing with. 

52.           HMRC accepted that it is clear from the JP Morgan case that the interpretation of a contract is a matter of law as to which expert evidence is not appropriate.  However, in their view Mr Kendall was at most dealing with mixed questions of fact and law (such as where and when a contract of insurance was concluded) and/or dealing with such matters as regards his opinion on the ECIC guidance.

53.           HMRC submitted that on the grounds put forward by the appellant much of Mr Johnson’s report should also be excluded.  They noted that Mr Kendall explained the framework for the regulation of insurance (following which he focussed on regulation and market practice in the UK) but that was also the case in Mr Johnson’s report.  Mr Johnson referred in his report to undisputed matters of EU legislation, such as insurance directives.  Moreover if any legal analysis on the regulatory side is irrelevant, much of Mr Johnson’s report would be irrelevant as well (noting, in particular, that Mr Johnson also relied on the ECIC).  He also referred extensively to the evidence set out in the witness statements and applied them in a regulatory context. 

54.           HMRC submitted that Mr Kendall should have an opportunity to explain the views and experience which led him to make the statement regarding the significance of concluding insurance contracts given that many of the witnesses of fact were questioned on that statement.  This forms part of his analysis, which lead to his overall conclusions.  HMRC did not accept that the statement was mere assertion.  Mr Kendall explained the background to it but again he could be cross-examined on that. 

55.           HMRC said that any issues regarding Mr Kendall’s experience or independence or as to his understanding or selection of the facts could be put to Mr Kendall in cross-examination.  The tribunal could then decide what weight they wish to put on Mr Kendall’s views.

56.           Finally, HMRC said it is difficult to see how the second report is in any way inadmissible given that it was a response to Mr Johnson’s report, which the appellant wished to rely on.

Conclusion on expert evidence

57.           Our view was that the majority of Mr Kendall’s report was inadmissible whether on the basis that it was legal opinion, which ought properly to be addressed by way of legal submission, or that it sought to address the very issues for the tribunal thereby usurping the tribunal’s proper function.  Whilst some of Mr Kendall’s comments were mere assertion, we did not consider that Mr Kendall’s treatment of the factual evidence raised sufficiently serious doubts on his impartiality and independence for his reports to be excluded on that ground. 

58.           It is the tribunal’s function to find the facts and apply the law to those facts.  In this case the legal question for the tribunal is whether, under the applicable VAT rules and case law, (1) as HMRC argued, Advantage made insurance supplies to UK customers through a UK FE, in the form of the appellant’s human and technical resources, and received the appellant’s broking, claims handling and underwriting support services at any such FE or (2) as the appellant argued, it made such supplies received the services at its BE in Gibraltar.  The appellant can only recover the input tax in dispute in that second case.  It did not appear to be disputed that the tribunal must apply the applicable VAT rules and case law on the basis of the tribunal’s findings on the effect of the contractual arrangements between the parties as construed under English law principles and on the operation of the arrangements in practice in the light of the witness and documentary evidence. 

59.           We can see that where the required fact finding and legal analysis takes place in the context of a particular industry which employs concepts, terms and practices specific to that industry, the view of a person with particular expertise and knowledge of those terms and practices, as generally applied in that industry, may, in some circumstances, assist the tribunal in finding the relevant facts and making its assessment.  Mr Kendall’s evidence, however, was largely not addressed to any such matters. 

60.           As regards the majority of the report Mr Kendall was not acting as a skilled witness of opinion, as regards any relevant practice within the motor insurance industry, or of fact who, as set out in Kennedy, collates and presents to the court in an efficient manner the knowledge of others in his or her field of expertise.  Mr Kendall simply gave his own views on the factual evidence and drew his own legal conclusions from that evidence and his own construction of the effect of the contractual arrangements between the parties.  In many places it was not entirely clear whether the opinion he then set out was expressed as a matter of general, regulatory or VAT law.  On the face of it, given the way the questions he was asked to address were phrased (being the very VAT questions the tribunal has to answer) and the manner in which he expressed his conclusions, in many sections he appeared to address the very legal question before the tribunal on the basis of his own view of the facts and law. 

61.           HMRC argued that in fact it was clear that he was analysing the position from a regulatory perspective on the basis of assumed facts. There were sections where Mr Kendall specifically commented on the CEIC but otherwise his conclusions were more broadly framed in response to the questions he was asked to address.  However, we would not regard matters of regulatory law (if relevant) as outside the scope of the competence of the tribunal.  We see no reason why any such legal issues should not be addressed by way of legal submissions in the usual way (and of course HMRC could act on the basis of Mr Kendall’s advice in doing so). In any event, we could not see that an analysis of the regulatory tests is relevant to the applicable VAT tests; they are two different tests under two different statutory regimes.   

62.           We decided, however, to admit the following limited sections of Mr Kendall’s reports.

(1) Mr Kendall’s comments on what he regarded as involved in, and the most significant aspects of, a motor insurance business (at paras 7, 27 and 28 of the first report).  Arguably this may be viewed as an opinion on industry practice which may potentially be of some assistance to the tribunal.  Moreover as the witnesses of fact were asked about this statement, we considered that it was appropriate for Mr Kendall to be given the opportunity to comment further on this aspect (and for him to be cross-examined on it).

(2)  Mr Kendall’s general outline of the insurance specific terms as he understood them and on the regulatory framework (in paras 11 to 26 of the first report).  This merely set out the legal and regulatory framework which was unobjectionable background and clarification of how Mr Kendall was using the relevant terms (albeit some of it may be redundant as we were not admitting the whole of the reports).

(3)  Mr Kendall’s views on the CEIC guidance (in paras 94 to 97 of the first report).  We considered that it is a matter of expert opinion how that guidance is applied in practice by the UK regulatory authority.  We did not consider it was clear whether that was what Mr Kendall was addressing in that section of his report but that could be clarified in his oral evidence.  For similar reasons we decided to admit paras 5, 22 and 23 of the second report where Mr Kendall set out a view on regulatory practice and looked at what Mr Johnson said in his report on the CEIC issue.

63.            We note that at the point in the hearing at which we decided on this issue it was not entirely clear how or to what extent the evidence on regulatory practice in the UK and Gibraltar was relevant to the parties’ arguments.  As noted we could not see that any analysis of the legal regulatory position was relevant.  HMRC’s skeleton argument contained a number of paragraphs relating to the CEIC guidance and we had not yet heard their full submissions.  We considered that the most effective way of dealing with this was to deal with the relevant evidence and decide on the weight, if any, to be attached to it having heard the full submissions.  In the event, in their submissions HMRC said that the CEIC guidance “is to be followed only so far as it matches the guidance given by the Court of Justice in its case law interpreting the concept of [FE] under the VAT Directive.”  We could not see that the guidance set out in the ECIC was relevant and neither party ultimately put any argument to the contrary.   We have not, therefore, attached any weight to that evidence.

Facts - Overview of the appellant and Advantage

Structure and business

64.           The appellant is a private company limited by shares incorporated in England and Wales on 20 October 1995 with company number 03116518.  During the Period, it was authorised in the UK by the Financial Services Authority and, from 1 April 2013, by the Financial Conduct Authority (“FCA”) to arrange deals in, assist in the administration and performance of, and make arrangements for insurance contracts with commercial and retail customers. 

65.           The appellant sells insurance on behalf of a panel of insurers which, during the Period, comprised about 20 insurers, including Advantage.  It sold insurance on behalf of Advantage pursuant to services agreements concluded with Advantage on 1 October 2005, 25 February 2010, 21 December 2012 and addendums to those agreements (the “service agreements”).  The appellant’s main offices in Bexhill were not dedicated to selling Advantage policies; panel business was also sold from Bexhill.

66.           Advantage is a private company limited by shares incorporated in Gibraltar and has at all times operated from business premises in Gibraltar.  It underwrites UK private and commercial motor and motorcycle insurance for distribution through insurance intermediaries.  It was set up in 2002 initially to underwrite motorcycle business which the appellant’s underwriter at that time, Inter Hannover Limited, did not wish to underwrite.

67.           Mr Charlton’s evidence was that, from a regulatory perspective, Gibraltar was chosen as the place to base Advantage due to the ease and speed with which an insurer can be set up and authorised.  When Advantage was set up, the regime in Gibraltar, regulated by the Financial Services Commission in Gibraltar (“FSC”), also had lower capital requirements than the regime regulated by the FSA in the UK.  Mr Charlton was aware that there were also corporation tax advantages: Gibraltar had a zero rate of tax initially that changed to 10% (other than in relation to investment income).

68.           Throughout the Period, Advantage was licensed by the FSC on the basis that its “mind and management” was situated in Gibraltar, to provide general insurance contracts in or from Gibraltar relating to land vehicles (excluding railway rolling stock) and to insure motor vehicle liability risks into the UK on a “passporting inward service” basis.  This permitted it to provide these insurance services from Gibraltar.  It was not licensed on the basis that it was supplying insurance in the UK through a branch in the UK.  Advantage’s annual insurance returns submitted to the FSC make it plain that Advantage was operating on this basis.  It appears that the FSC has not raised any concern on Advantage operating on that basis nor have the UK authorities.

69.           Mr Charlton said that one of the advantages of Advantage being based in Gibraltar was that the regulator was “literally around the corner”.  During his time as CEO of Advantage, he had a good relationship with the regulator and was in frequent contact to “make sure that things were going correctly”.  On Mr Godfrey’s evidence, the good and “open” relationship between Advantage and the regulator continued following Mr Charlton’s retirement.

70.           The main services of insurance broking, claims handling and underwriting support and data analysis have always been outsourced by Advantage to the appellant although claims handling was on occasion further sub-contracted to another company, Endsleigh.  Before 2011 Advantage placed business through other intermediaries who acted as sub-agents of the appellant.  All intermediaries were aware that Advantage was the insurer and the policies issued to customers set this out.

Ownership of the appellant and Advantage

71.           From September 2006 until 3 February 2009, the appellant and Advantage were both indirect, wholly-owned subsidiaries of IAG International Pty Limited (“IAG”).  In February 2007, IAG purchased Equity Syndicate Management Limited (t/a Equity Red Star).  Mr Charlton was at that time a joint active underwriter at this entity.  His appointment as CEO and director of Advantage shortly after this in July 2007 was aimed at improving Advantage’s business which had been struggling in the market.

72.           In February 2009, a number of the managers of Advantage, including Mr Charlton, purchased Advantage through a management buy-out.  At that time, therefore, the appellant and Advantage did not share a common parent company but they had almost identical ultimate shareholders of their respective parent companies.  They were not identical as the appellant’s employees held an 18.1% shareholding in it.  The appellant and Advantage were “related parties” for the purposes of UK Generally Accepted Accounting Practice as a result of the commonality in the shareholders.

73.           Hastings Insurance Group Limited (“HIGL”) was formed in 2011 as the holding company of the appellant.  On 17 April 2012 HIGL purchased Advantage, following which the appellant and Advantage were again part of the same corporate group. 

74.           Mr Charlton retired as CEO in January 2014 and Mr Godfrey took over the day to day management as Advantage’s managing director.  From April 2013, Mr Charlton’s role as CEO of Advantage was to expand Mr Godfrey’s knowledge base into finance and other areas that were necessary for him to oversee in order to take over responsibility for running Advantage.

75.           Throughout the Period, the appellant and Advantage had separate management structures and had no mutual directors.

Separation of Advantage and the appellant.

76.           At the time of the buy-out in 2009, a decision was taken to strengthen Advantage’s corporate governance.  Mr Nigel Feetham, a lawyer who is a local Gibraltar resident, was appointed as a non-executive director and chairman and additional staff were recruited.  Three personnel, including Mr Godfrey, who were directors of Advantage but also worked for the appellant, resigned as directors.  Advantage also introduced a formal audit committee which reported to the board. 

77.           Advantage’s staff levels increased such that by December 2013 Advantage had 19 full time equivalent employees.  These included actuaries (Mr Richard Kelsey as a consultant actuary from 2010/11 and Mr Pablo Morales as senior actuary from August 2011 to December 2014), Mr Godfrey, two underwriters, a claims team comprising a director and six others, a finance director, four finance staff and several administrative staff.

78.           Mr Charlton acknowledged that Advantage and the appellant were very closely linked but he said that what was “very important to us as a group at the time of the buyout in 2009, was to keep the management [of them] separate”.  When he became CEO of Advantage, it became fairly apparent that Advantage was losing significant amounts of money and one of the reasons was that the “Hastings marketing arm had too much power and influence over what was happening in pricing”.

79.           Mr Charlton said that, prior to the sale of Advantage to IAG, the owner of the Advantage business was running it as his own, and there were a number of things in the rating structure of Advantage which put it at a disadvantage but allowed the appellant to grow business.  That had to be stopped.  Steps were taken so that “[the relevant employee at Hastings] was in charge of his part of the business, Hastings”, and Mr Charlton was in charge of Advantage, and “we would manage our own businesses to make profit, and their own success.  And that was a key factor…… we did not put these structures in place because we wanted to claim back VAT; we put these structures into place for a real commercial reason.....to get rid of that conflict between the two parts”.

80.           Mr Charlton said that whilst the two businesses were closely linked “particularly on a day-to-day basis…..you do not manage a business looking at your corporate structure.  You manage your business to be the best for the unit you are looking after.” He said that Advantage “is not there to make Hastings more competitive.  It is there to make money for Advantage”. The two businesses were “not integrated in the sense that day-to-day management of the business is the same.  Far from it.”

81.           Mr Charlton continued that the group was “not the same as a Direct Line or an Esure” in that “we have, on purpose, split the two businesses apart to take away that conflict of interest and irritation between the marketing arm and the technical side of the business, the underwriting side”.  He said that, by now, one could not operate without the other but he thought that either of them could obtain what the other provided from elsewhere.  He thought that the model gave them some unique selling points in that “both of them were doing their best….for themselves, which then builds a better business for the group”.

82.           He said that obviously the people employed by the appellant have a very good knowledge of what Advantage does.  When Mr Godfrey moved from the appellant to Advantage in April 2013, one of the reasons for the fairly lengthy handover period, was to ensure that he took on board the need to manage Advantage for Advantage’s sake, as opposed to being influenced by what happened at the appellant.

 Key elements of insurance business

83.            It is clear from the witness evidence that an insurance business of this kind operates as follows:

(1) The insurer contracts with a customer to provide insurance cover for a specified period according to the terms and conditions set out in the insurance documents (such as a policy and insurance certificate) in return for an agreed premium. 

(2) On receipt of a claim under the insurance policy, the insurer handles the claim and arranges payments.  Required action may include arranging repairs, appointing loss adjusters, dealing with third party claims against the customer and instructing solicitors.

(3) To provide this service the insurer requires a “front office” or marketing function which deals with the provision of quotes to customers for insurance cover, the sale of policies to customers, the issue of the required policy documentation, the day to day claims management as regards the customer and customer complaints as regards these services. 

(4) To provide the insurance product the insurer determines the terms of the insurance policy, decides what risks it is prepared to cover and what price it is prepared to accept for doing so. 

(5) As insurance is a regulated industry, the insurer has to ensure that all regulatory and compliance conditions are met.  The insurer has to maintain adequate capital to support its operations, which involves, among other matters, a valuation of reserves required to meet claims.

(6) The business requires a finance function to manage the general day-to-day running of the finance, the receipt of premiums, the payment of claims and the maintenance of accounting records.

(7) The insurer will typically need to have an asset management and treasury function, as insurance companies generally receive premiums some time before any claims are required to be paid.

(8) The insurer may decide to “lay off” some of its risk to reinsurance companies, in which case it will need to negotiate the reinsurance terms and deal with claims notifications to the reinsurers. 

84.           As noted, Mr Kendall said that from a commercial perspective the most significant activities involved in the conduct of insurance business are “the effecting of contracts of insurance and the carrying out of those contracts”. At the hearing he said he formed that view as he thought that a person cannot be regarded as having an insurance business unless it enters into contracts of insurance and provides the relevant services under those contracts.  It may have assets to manage and it may employ people who have expertise in insurance, but that does not mean that it is carrying on insurance as a business. 

85.           We do not consider that any weight can be attached to these comments in terms of assessing the functions of this business from a VAT perspective.  It was not clear whether Mr Kendall meant this was the view of the regulator or, if this was an industry view, or to be applied some other context.  Mr Kendall did not put forward any basis for how he was measuring the significance or value of different functions involved in an insurance business. 

86.           Plainly an insurer could not function without the ability to sell its insurance products and to carry out its obligations under the insurance contracts sold.  However, it is also clear from the nature of an insurance business, that the insurer would not be able to carry on a successful insurance business without the ability to assess what risks to insure for what price and the means to cover claims made under the policies.  It does not require specialist knowledge to realise that the success of an insurance business must depend on the ability to balance the premium charged against the likely realisation of and amount of claims.

87.           We found the evidence of the witnesses of fact more useful in understanding what is involved in an insurance business.  Mr Charlton described an insurance business as involving a number of key functions.  Part of it is, what he described as, the mechanical or logistical functions of giving quotes, accepting quotes, issuing policies, dealing with the policy administration and with any subsequent claims that arise from that policy.  In his view, the key feature, however, is the ability to understand what risks and liabilities are being assumed.  The insurer has to understand what type of cover is being provided, the breath of cover and what the particular risk is bringing to the insurance book.  That requires technical pricing and an understanding of past risk results from those insured. In insuring a driver, for example, there are a large number of questions that are asked to assess the risk and which are factored into the “risk price”, the net premium, the insurer receives for the insurance. 

88.           He considered that it is particularly significant that when the price is quoted to the customer it is not known what the actual cost of that policy will be for the insurer for quite some time.  The insurance cover lasts typically for twelve months but claims may arise some while after that; large claims can take years to settle.  A big part of what an insurance company does is to analyse the patterns of past performance to understand the likely future performance of the various risks.  A major element of that is, first of all, deciding what to insure.  Typically, Advantage only quotes for about half the possible risks it could quote for and then actually accepts a much smaller proportion of that half.

89.           He continued to explain that once the insurer has taken the risk on board there is the question of how claims are dealt with, the claims philosophy (which is set by Advantage) and what reserving levels the insurer wishes to establish, so that it can anticipate the likely claims as well as it possibly can.  A lot of the analysis on that is done by technical staff.  But what comes out of that analysis is the need to make decisions on what rates to charge, what risks to accept, and the need to monitor the results of what has been done.  A key factor is to measure what actually happens and make amendments to the insurance business based on that.

90.           He said that the insurer also requires the ability to run an investment portfolio, finance and reinsurance to limit its liabilities and a good understanding of what risks it wishes to take.  He said that “all goes into the melting pot of deciding what you need to protect yourself and make money.”  From a commercial perspective, the most significant factor is “to make sure that the business has sufficient premium to pay its liabilities and hopefully make a profit”.

91.           Mr Charlton acknowledged that the appellant carried out functions which are part of an insurance operation, in effecting contracts of insurance and carrying out those contracts, receiving premiums and handling claims.  He said that what the appellant did was only a small part of an insurance business as he had described it above.  He explained that, looking at how the appellant sells insurance and “how any insurance intermediary sells business, typically it is….all system driven”.  The quotes are given by a call centre member of staff, who inputs information into a computer system and “out pops the answer, which is the premium”.  If the customer likes that answer they accept the policy and pay the premium, which is then collected by the call centre operator and the policy is issued by a printing system:  “Any claims are then rung in, subsequent claims.  Cover is checked to make sure they have got the cover, and a claim system will deal with the technical, the logistics, of actually doing that”.  However, as noted, in Mr Charlton’s view:

“If it was charging the wrong premium, if it was not having sufficient capital and solvency to be a regulated business, if it was not analysing the mix of business it was bringing in, understanding the risks that they brought, they would be out of business extremely quickly.”

Terms of the services agreements

92.           The terms of business of the services agreements were negotiated between Mr Tobias Van De Meer, as managing director of the appellant, and Ms Amanda Truelove and Mr Charlton on behalf of Advantage.  We accept on the evidence set out below that, although the agreements were between related companies, the parties negotiated their terms with a view to their own commercial interests as though they were unconnected or on an arm’s length basis.

93.           Mr Charlton said:

“we both ensured we had a commercial arm’s length agreement in relation to commission paid to Hastings for the services received by Advantage.  I wanted to ensure Advantage remained independent, primarily because a key market differentiator for Advantage and Hastings was that the independence of Advantage decision making reduced the pressure for sales considerations to be taken into account in underwriting decisions and also because this was insisted upon by the FSC in Gibraltar”.

94.           From the board minutes in the bundles, the agreements were considered separately by the boards of the appellant and Advantage.  Advantage’s board meeting minutes dated 6 April 2009 stated, in relation to the 2010 service agreement, that “Hastings would be treated as an arm’s length service provider”. The appellant’s board minutes dated 30 July 2009 referred to it having reviewed the services agreement with a view to the arrangements being on an arm’s length basis being “clearly established”.  The arm’s length basis of the relationship under the services agreements was expressed in terms in the 2012 services agreement.

95.           Consistently with the contractual position and the parties’ intention to operate independently, the insurance policy documents explained to customers that Advantage and the appellant “act totally independently in the day to day running” of the businesses (as shown in the example car insurance policy in the bundles).

96.           Mr Johnson said the services were functions that are commonly outsourced.  Mr Kendall said the appellant’s activities are typical of those performed by an insurance intermediary.  Mr Charlton agreed the services were broadly normal in the market place.

Terms of the services agreements

97.           There were changes to terms of the agreements over the Period but the main terms, as set out in the services agreement dated 3 February 2010, were substantially the same through the Period.  In summary:

(1) Advantage appointed the appellant to provide underwriting, claims handling and support services in return for commission “for the purposes of enabling Advantage to undertake insurance business on a cross border services basis”.  The appointment was stated to continue indefinitely thereafter until terminated under the provisions of the agreement. 

(2) The appellant’s authority was confined to acting as (a) service provider to Advantage (b) as intermediary in relation to the arrangement of contracts of insurance and (c) as “agent of Advantage but not, under any circumstances, as principal” and it was to “have no liability or obligations for any Risks”.

(3)  The appellant was not required to arrange for Advantage the reinsurance of any risks but it could obtain and provide Advantage with any quotations for such reinsurance.

(4) Advantage was entitled to appoint or employ any other party to perform the Services and the appellant was not precluded from accepting similar appointments from any other company or companies.

(5)  The underwriting services comprised the provision of an underwriting, account monitoring and analysis service to include, acceptance of new risks, mid-term adjustments, renewals, and cancellations (within agreed parameters), rate modelling and setting recommendations for approval and sign-off by Advantage and the audit of brokers. This included that:

(a)    in accordance with instructions from Advantage from time to time, the appellant had authority to execute insurance contracts within the acceptance terms of the Advantage underwriting guidelines and its computer system;

(b)   the appellant was required to carry out pricing of products within the terms of the Advantage rate sets as may be amended from time to time by Advantage; and

(c)    the appellant was required to provide statistical analysis and reporting on a monthly basis for the purposes of account monitoring.

(6) The appellant had no authority to accept risks or amend products beyond the levels of authority specified above.  Matters which were outside the scope of the appellant’s authority under the agreement were to be “referred to Advantage for decisions as and when the matter arises”.  All other matters were to be discussed and reviewed at an operational review meeting (“ORM”).

(7) The claims service required the appellant to provide “a fully outsourced claims handling and management service” as set out in the appellant’s claims procedure manuals, as specified in the claims handling guidelines and as approved by Advantage from time to time.  This included authority:

(a)     to reject, adjust, compromise, admit, settle and pay claims and pursue recoveries for the account of Advantage in respect of risks (including by taking and defending legal proceedings), up to a maximum of £50,000 for each claim, within such guidelines as may from time to time be notified by Advantage to the appellant.  The specified threshold for the purpose of this provision was £25,000 in the previous agreement and was increased to £75,000 in the later 2012 agreement.  We refer to claims in excess of these thresholds as “large loss claims”;

(b)   to deal with salvage and recoveries including recoveries from any outward reassurance in connection with contracts;

(c)    to accept service of writs and claims, notices and other legal processes in connection with disputes and claims in connection with contracts;

(d)   to handle any claims in excess of the specified threshold under instruction from Advantage and to assist Advantage in the pursuit of resolution of such claims provided always that Advantage could assume the handling and conduct of any or all claims at any time; and

(e)    to appoint loss adjusters, average adjusters, surveyors, assessors and legal and other professional advisers under instructions from Advantage for the purpose of dealing with claims made under contracts.

(8) The appellant was required to submit a suite of reports, the format and content of which was to be agreed from time to time by the parties, which were to be presented at the ORM where any matters arising were to be discussed and addressed.

(9) Any change in the information provided or the process for managing the claims handling service was to be agreed between Advantage’s and the appellant’s claims directors and/or the Advantage managing director.

(10) The support services included the following:

(a)    The hosting and development of the IS2000 system upon which contracts were priced and administered.  The service levels were to be as agreed between the parties from time to time and any and all matters arising in connection with the IT services and support were to be discussed and reviewed at the ORM.

(b)    The provision of services for the collection of premiums from brokers in relation to contracts sold by them.

(c)    The provision of actuarial resources and support to enable Advantage to carry out loss reserving.  The appellant was required to provide support “on an ad hoc basis as and when required”.

(d)   Any and all matters arising in connection with the finance services and support were to be discussed and reviewed at the ORM.

(11) The appellant was authorised on behalf of, and acting as agent for, Advantage, to receive and collect premiums and other sums due on contracts on behalf of Advantage and to hold claims moneys.  Liability for risks was stated to pass to Advantage upon completion of a valid contract by the appellant or a broker and Advantage was to acquire title to all net premiums at the point they were paid by an insured to either the appellant or a broker.

(12) In consideration of the commission payable the appellant agreed to:

(a)     maintain proper books, records and accounts as were necessary to indicate the business underwritten and to enable Advantage to accurately identify and quantify its liabilities and retain all such records for so long as legally required to do so; and

(b)    give reasonable access to Advantage and its duly authorised representatives during normal business hours to all relevant records and accounts;

(c)     prepare accurate declarations in respect of each and every applicant for insurance where necessary;

(d)   provide the wording of slips, policies, certificates of insurance and endorsements in respect of contracts issued and to despatch such policies, certificates and addenda within such guidelines or instructions as may from time to time be notified by Advantage to the appellant;

(e)     collect and pay to Advantage, on a monthly basis, the net premium due in respect of policies written under the agreement in respect of contracts;

(f)     immediately advise Advantage of any claims or potential claims in excess of the specified threshold and provide Advantage with full claim details with the notification;

(g)    make available to Advantage each day an extract of customer and risk information in respect of all premium transactions for contracts including such statistical information as required by Advantage;

(h)   comply with the provisions of the Data Protection Act 1998 as data processor when dealing with applicants and entering data on to the Database.

(13) The appellant was entitled to take the commission due in respect of each contract it sold at the point at which cleared funds were received from the insured.  The difference between the premium and commission due in respect of each contract sold was to be transferred into a separate bank account, which was owned and operated by the appellant.  All sums due to Advantage were to be paid from that account and any and all interest earned on the balance held in that account from time to time belonged to the appellant.

(14) Advantage was required to pay all levies and other similar payments to the Motor Insurers Bureau, Financial Services Compensation Scheme, Motor Insurers Anti-Fraud and Theft Register and Thatcham in respect of contracts entered into under the agreement.

(15) The appellant was required each calendar month to calculate all insurance premium tax (“IPT”) to be added to the gross premium and provide the relevant figures to Advantage.  Advantage was required to complete all returns and make all payments of IPT to HMRC.

(16) The appellant was required to collect all income arising from the contracts and to submit to Advantage monthly in arrears within 20 days of the close of the month end a breakdown of (a) all accounts and bordereaux providing details of contracts in the relevant month; and (b) the amounts of premiums relating to contracts showing the commission payable in respect of the relevant month and any claims paid. Settlement of amounts due to Advantage were to be made within 90 days of the end of the relevant month.

(17) An ORM was to be held at least once in each calendar month at which all matters relating to the appointment were to be reviewed and discussed.  The ORM was normally to be held by videoconference between representatives of the appellant (including the underwriting, claims and insurer and product development directors) and representatives of Advantage (including the CEO and claims and finance directors). 

(18) Advantage acknowledged that the ownership of the database was the property of the appellant.  This was defined as the full customer database in relation to the insurance business and as including all relevant fields, field headings, and interpretative and supporting tables; and all records including records to current insureds, past insureds (from the date of commencement of the contract), any additional drivers under a policy of motor insurance, any prospects to whom a quote was supplied but where no policy of insurance was issued (in the two years prior to the date of the agreement).

(19) The appellant was required to indemnify and hold Advantage harmless against all claims, losses, outgoings and liabilities which Advantage may suffer or incur by reason of any breach by it of its obligations pursuant to the terms of the agreement.

(20) Either party could terminate the agreement without cause subject to giving a minimum of six months prior notice in writing to the other party.  There were also provisions for termination in the event of certain breaches.

98.           The agreement also set out risk criteria as regards the insurance cover which Advantage was prepared to cover for private cars and motorcycles.  These included a number of instances where the risk was to be referred to the appellant’s underwriting director or Advantage:

(1) Referrals had to be made to the underwriting director/managing director of the appellant, as regards car insurance, for cover for (a) drivers outside specified age parameters and (b) for vehicles worth over £50,000 for new car business and £75,000 for existing car business and (c) for renewals, where the claims exceeded £100,000.  Insurance cover for cars worth more than £100,000 had to be referred to Advantage and for renewals where the claims cost exceeded £500,000.

(2)  Referrals had to be made to the underwriting director/managing director of the appellant, as regards cover for motorcycles worth more than £20,000 for new business and for renewals for risks exceeding £100,000.  Cover for motorcycles worth over £30,000 and for renewals for risks exceeding £500,000 had to be referred to Advantage.

99.           The document also included a copy of claims handling guidelines, the main provisions of which were as follows: 

(1) It was noted that there were a number of manuals, including the appellant’s claims procedure manuals which were stated to be approved by Advantage as forming “the basis of acceptable practice”.  Advantage authorisation had to be obtained prior to any significant/material amendments being made to the operating manuals.  Advantage retained the right to assume handling and conduct of any or all claims at any time.

(2) Advantage expected that all insureds would receive full indemnity of their losses as promised within their policies. This should happen within a reasonable timeframe consistent with the insureds’ expectations.

(3) It was stated that Advantage would conduct audits of the appellant’s claim functions to ensure that claims were being handled to an acceptable standard, with minimal financial leakage, and that claims procedures were adequate. Audits were to be undertaken pursuant to the Advantage audit schedule. The appellant was required to respond to comments raised in these reports within 10 working days.

(4) Advantage said it would endeavour to audit at least 1% of weekly claim payments and would audit all files with incurred movements in excess of £25,000.  The appellant was to respond to comments raised within Advantage audit reports within 10 working days.

(5) Advantage instructed the appellant to maintain appropriate controls, systems and procedures to ensure that the provision for claims outstanding was, at all times, sufficient to cover any liabilities that have been incurred on insurance contracts “as far as can be reasonably foreseen”.

(6) All claims matters relating to reinsurance had to be referred to Advantage, as Advantage was responsible for reporting reinsurance claims directly to their reinsurers.  The appellant was required to provide every assistance in this regard.

(7) Advantage instructed the appellant to maintain appropriate controls, systems and procedures to ensure that the provision for claims recovery was, at all times, accurate to represent a realistic recovery prospect of outstanding debt incurred on insurance contracts as far as can be reasonably foreseen.

(8) The appellant was required to ensure that any proceedings served or issued directly on or by the Advantage, were issued in the name of Advantage. Advantage confirmed that the appellant was an authorised signatory on all legal documents on Advantage’s behalf.

(9) As regards large loss claims:

(a)    The appellant was required immediately to advise Advantage of any large loss claims or potential large loss claims and to provide full details with the notification.

(b)   The appellant was required to investigate and negotiate large loss claims as appropriate, as follows; the appellant was to maintain adequate systems either electronically or manually to ensure that large loss claims were identified at the earliest possible opportunity, all documents and correspondence, including e-mails, were to be suitably imaged and indexed to enable full consideration to be made by Advantage, immediately the claim was so identified Advantage had to be notified, via e-mail, IS2000 or electronic bordereau, within 5 working days the appellant was required to provide advices to include a summary of claim, liability recommendations, relevant documentation, reserve recommendations and the plan/tactics. 

(c)    Advantage intended that following its agreement to such recommendations, the appellant would be able to proceed generally without having to refer back on minor points

(d)    Significant developments had to be notified to Advantage such as, but not limited to, liability, settlement/quantum, medical reports/rehab INA Reports, expert reports, such as opinion from counsel, reports from solicitors and investigators, writs and pleadings and the settlement of claim.

(e)    The appellant had to notify Advantage via an appropriate method to include specified details. 

(10) Any movement beyond £5,000 to the authorised incurred total had to be notified to Advantage with an appropriate reserve breakdown.

(11) In the event the appellant recommended entering a defence of fraud, on cases in excess of £50,000 or on linked cases (Fraud Rings) where the total combined value was in excess of £50,000, such cases had to be referred to Advantage for authority prior to proceeding.

(12) The appellant had to ensure that the reserve was reviewed in line with Advantage’s reserve philosophy and amended as appropriate upon payments being made.

(13) All Accidental Damage aspects, subject to indemnity, could be handled by the appellant without referral to Advantage.

(14) The appellant was required to notify Advantage in advance of any payments being made in excess of £250,000

(15) Cases under such instruction must not be either closed or reduced below the £50,000 threshold without referral to Advantage.

(16) For the avoidance of doubt, any matters where authority had not been explicitly granted to the appellant in the above guidelines had to be referred to Advantage in the first instance.

(17) The Large Loss Committee (“LLC”) was to consider all reserve movements in excess of £100,000.  The LLC was to meet once a week.

100.       As regards the appointment of loss adjusters and other professional advisors for the purpose of dealing with claims made under the contracts:

(1) The appellant was required to maintain a panel of competent suppliers of services as necessary.

(2) Such suppliers were to be audited by the appellant as appropriate and audit reports were to be provided to Advantage.

(3) Advantage was to be updated with a list of suppliers who generated fees in excess of £50,000 per annum.

(4) Advantage was to be updated with a list of suppliers who generated claim payments in excess of £100,000.00 per annum and continuing Ml capturing the actual spend.

(5) Advantage was to be provided with copies of all sub-contract agreements.

(6) Advantage retained the right to object to the appointment or request the removal of any supplier.

101.       As regards its liabilities to maintain books, records and accounts the appellant was required to provide  a number of management information reports to Advantage via e-mail including:

(1) on a monthly basis, reports on open claims, recoveries, large loss analysis, a claims scorecard, claims statistics, settled claims, complaints, the appellant’s staff headcount (including outstanding workloads), theft investigation & repudiation, class of business code listing, a pack of the appellant’s internal & external claims audit reports;

(2) on a weekly basis “+£25k Report” and claims paid bordereau; and the quarterly authorisation limits of the appellant’s claims staff.

102.       All claims matters relating to reinsurance had to be referred to Advantage, as it was responsible for reporting reinsurance claims directly to their reinsurers.  The appellant was required to provide “every assistance”.  All reinsurance first advices had to be referred to Advantage and accompanied by an appropriate “Reinsurance First Advice”. Significant developments had to be reported to Advantage and accompanied by an appropriate “Reinsurance Update”.

103.        Claims of the following nature had to be referred to Advantage, regardless of the level of cost incurred; intimation of a claim that involves a continuing regular payment (or a periodic payment under the Courts Act 2003), fatal injury to a third party where such party leaves a spouse or issue or other dependant(s);  brain injury; spinal injury resulting in partial or total paralysis of upper or lower limbs; amputation, or loss of use of, upper or lower limbs;  permanent disability rating of 50% or more.

104.       The appellant was required to arrange regular meetings with Advantage, either in person, via telephone or via video link, to discuss claims matters generally and provide relevant updates in respect of any claims handling issues.

105.       In the event of Advantage needing to escalate any issue or complaint, this was to be referred in the first instance to the appellant’s operations manager. If further escalation was required then this was to be made to the appellant’s claims director.

Changes to the agreements

106.       As regards changes to the agreements, Mr Van De Meer or Mr Pavey generally made suggestions to Ms Truelove and/or Mr Charlton.  Mr Charlton said he regularly challenged these proposals and that these were real negotiations even though they took place on an intra group basis.  In carrying out these negotiations, Advantage considered the quality of the appellant’s services, and took into account information provided at the monthly ORM about:

(1) Whether the appellant provided customers with a good service.

(2) Whether the appellant answered calls to the call centre quickly.

(3) Whether the appellant was accurate when performing claims handling and underwriting services (for example, whether it provided potential customers with the right quotes, within a proper time frame).

(4) Whether the appellant was complying with the services agreement.

(5) What protection the appellant could give Advantage’s business (for example, by the appellant making sure its system does not prefer other panel insurers over Advantage).

(6) The total cost of the appellant’s services relative to the reasonable cost of Advantage carrying out the same services. Advantage examined commission levels applied by brokers in the marketplace other than the appellant (as Mr Charlton had broker contacts who he could speak to).

Re-broking agreement

107.       The services agreement in force in 2009 was amended on 29 January 2009 with an endorsement that was stated to be effective from 3 February 2009 and to be in force for 5 years.  Under this document the appellant agreed not to re-broke with other panel insurers agreed parts of Advantage’s renewal business in return for a one off payment of £6 million and subject to the following main terms and conditions:

(1)  “Advantage will identify and agree with Hastings on a regular basis which types of business Hastings will not re-broke to an alternative Insurer, the non-rebroke scheme”. 

(2) This “non-broke renewal scheme will generate for Advantage up to £60,000,000 of the current Gross Written Premium over the first 36 months of this scheme”.

(3) “Advantage will ensure that the renewal rate for each case will not be more that 2% above the underlying rate increase within Hastings book of business”.

(4) “If the retention rate falls below 65% on the non-rebroke scheme Hastings will have the right to negotiate the renewal premium rate reduction up to 5% and Advantage will have the right to negotiate a reduction in any renewal fees where it is in the joint interests of both business”.

(5) “Advantage will pay the standard commission rate for renewals under the non-rebroke scheme and in addition will pay a one off sum of £6,000,000 to Hastings for this scheme.”

(6) “Advantage will not have the right to offer any business within the non-rebroke scheme to any other broker or intermediary or to deal direct with the customer”.

(7)  “In the event of the sale of Advantage this endorsement ceases and Hastings have full rights to all the customers under the non-rebroke scheme”.

(8) “Any annual review will be held in February to ensure that the arrangement is on track to generate the required premium for Advantage”. In fact the arrangement terminated on or around March 2010 because the gross written premium target had already been achieved by then.

108.       Mr Charlton explained that Advantage decided to enter into this agreement because Advantage was losing money at the time and he felt that it was commercially sensible to pay the money to preserve this profitable business with the appellant: “to protect Advantage I felt it was worth paying the money [£6 million] across to Hastings to keep the profitable business with Advantage and Hastings agreed.”

109.       Mr Pavey said that, at the time, Advantage was quite concerned about their renewal book.  He explained that loss ratios for renewal business, being the ratio of claims expense to premium income, are traditionally lower than loss ratios for new business.  He said that Advantage were very keen to protect their book so they wanted the appellant to ensure that they wrote at least a target amount of renewal premiums.  In cross examination he confirmed that the appellant did not have such agreements with other insurers but he thought that “that sort of agreement is very common place within the insurance sector”.  Mr Kendall confirmed that he had seen other instances of similar agreements.

110.       Mr Godfrey said that he thought that generally the appellant would not re-broke to a different insurer unless a customer came back with a price query.  He noted that at this particular time there was a change of the system that the appellant was implementing, “so some of this business was being converted from one system to another.  When it went onto the new system it went on as a new policy.  So there was more chance at that stage that it could have been replaced”.

111.       He was asked why Advantage paid £6 million to the appellant for the agreement if the appellant would not have re-broked in any event unless queried by the customer.  He said “I guess there was nothing to stop Hastings re-broking it if they chose to re-broke it” and referred again to the fact there was a change of system that was going on as well, “that may have led to some business being replaced”.  It was put to him that the only logical explanation was that it did not matter that Advantage paid £6 million for a service which the appellant would have been prepared to offer for nothing because of the commonality of ownership. He said again “it would have protected ….the renewal book, which is generally a more profitable business for an insurer” and whilst he was not close to it at the time, he thought that the reason for the agreement was “to get that certainty over it”.

112.       Mr Charlton confirmed that Advantage was not aware whether the appellant had entered into non-rebroke agreements with other panel insurers.  We accept that the agreement was entered into for commercial reasons and was of a type which may be entered into between unrelated parties.

Areas not conducted in accordance with the agreement

113.       Mr Godfrey explained that there are a few matters which either are not governed by the services agreement or were handled in a different manner to that set out in the agreement.  IPT returns were compiled and submitted by the appellant on Advantage’s behalf although payments of IPT were made by Advantage.  Mr Lee had an informal role in relation to negotiating reinsurance contracts as set out below and the reserve process is different as also explained below.

Commission

The appellant’s commission

114.       Under the commission arrangements, as set out in the services agreement and explained by Mr Godfrey, the appellant was entitled to the following commission:

(1) It received “base commission” which was the difference between the gross premium (the amount charged to the customer (plus IPT)) and the net premium (the risk price set by Advantage).  The appellant itself set the gross premium, thereby setting the level of its own commission, as explained below.

(2) It was entitled to an uplift in the base commission if it transpired that it was an insufficient return for the services provided for the purposes of UK transfer pricing rules.  At the time of the hearing it had not been necessary to pay any such additional amount. 

(3) From 2012 onwards, the appellant was entitled to performance commission, which was paid on a monthly basis, if certain agreed performance targets were met.  This comprised two elements: 

(a)    “Anti-Fraud Over-rider Commission” was payable by reference to how well Advantage’s loss ratio performed against a target ratio.  Mr Lee described this as reflecting the benefit that Advantage gained from the appellant’s anti-fraud work through improved loss ratios and profit share from Advantage’s reinsurers; and

(b)   “Claims Handling Commission” was payable on the appellant meeting certain targets or key performance indicators (“KPIs”) in relation to the delivery of its claims handling service.

115.       Mr Pavey explained that the loss ratio target and the KPIs were set on a yearly basis by agreement between the parties.  He was responsible for negotiating the terms of the performance commission with Advantage.  He discussed the proposals internally with Mr Lee who inputted into the process as regards the target for the claims handling commission.  The process started with Mr Godfrey giving him the recommended targets which were set by reference to historical data such as the rate for attrition in the appellant’s claims department, Advantage’s net premium forecasts and its budget. 

116.       Mr Godfrey said that Advantage used its knowledge of expense ratios in the market to set the KPIs.  The ratios were made up of sales, acquisition costs and expenses of running the business, such as staff costs and IT, as a proportion of gross written premiums.  Mr Godfrey and Mr Lee noted that adjustments could be made during the year if the parties so agreed. 

117.       During the year, a number of re-forecasts may be carried out according to, for example, how the premium was being written and earned by Advantage and how the appellant was performing against its claims handling targets.  If the appellant and Advantage thought that it might not be possible to achieve all the KPIs the amount that Advantage paid monthly was adjusted to reflect that.  The final amount paid for the year was finalised during the year end process, when the final review of performance against the targets was completed.

Handling payments

118.       Mr Pavey said that the appellant initially received customer payments into a UK clearing account.  On a daily basis it moved the amount of the net premium and IPT due to the insurers into one of two UK accounts (one designated to Advantage and the other to the rest of the insurers).  The amount of the income which represented the appellant’s commission was transferred to a separate broking account.  The insurer was informed, by class of business, of the gross premiums charged to customers, the amount of IPT, the base commission earned by the appellant and the amount due to the insurer. 

119.       The appellant paid the net premium and IPT amounts due to Advantage to its bank account in Gibraltar.  Advantage was paid these amounts following receipt from the customer.  Other panel insurers were paid these amounts within a fixed period of inception of the insurance policy regardless of when the customers paid the premium.  Mr Pavey noted that, in practice, payments due to Advantage were paid earlier than provided for in the services agreement.  Earlier payment was made because this had treasury advantages and the practice was agreed with Advantage to enable it to prepare for the introduction of Solvency II (as the balance owed was subject to a counterparty risk charge).

120.       Mr Pavey explained that customers’ policies were “time on risk”; they were only covered by the relevant insurance for the length of time the insurance policy was live.  If a customer fell into arrears with his premium, the appellant initiated a “chase cycle” to try to get the customer back on track with the payments but, if that failed, the appellant would cancel the policy.  Where an amount due from the customer for the time on risk was not recovered, the appellant still had to pay the net premium to the insurer.

121.       The appellant had a separate bank account from which it made payments to customers for claims handled on behalf of Advantage.  The appellant provided Advantage with a weekly file which showed movements in the account.  The appellant estimated how much it would spend in paying out claims each week and Advantage made several payments to the appellant over the course of the week to cover the amounts.  If the appellant had to make a large pay out on a claim (typically above approximately £100,000), Advantage was asked to transfer the appropriate amount prior to payment being made by the appellant.

122.       When the appellant undertook claims handling on behalf of insurers other than Advantage, the process for the funding of payments for claims was similar. The principal differences were that payments were made by other insurers typically on a monthly basis (as the amounts involved were much smaller than with Advantage) and any payments over £10,000 were required to be paid up front.

Overview of the appellant’s income

123.       Mr Pavey said that the appellant’s commission earned from policies underwritten by Advantage has declined as a proportion of the appellant’s total turnover: for the year to 30 June 2009 such commission comprised 47.3% of total turnover whereas, for the year to 31 December 2013, it was 26.3% of total turnover.

124.       In 2009 approximately 90% of car insurance policies transacted by the appellant as an intermediary were made on behalf of Advantage.  This proportion has fluctuated but, in the year ended 31 December 2013, Advantage underwrote 90% of the insurance policies which the appellant arranged for new customers.  All policy sales were made to UK customers.

125.       The business has expanded away from motor car and motorbike insurance to other personal lines of general insurance including van and house insurance (although that did not start until 2015).  At the end of 2013 private car insurance still accounted for 89% of the live policies sold by the appellant.

Provision of services to other insurers

126.       Mr Pavey said that the work undertaken by the appellant’s broking team in selling insurance policies was generally the same for Advantage as for other panel insurers.  Otherwise the role undertaken for other insurers was more limited.  As the appellant did not undertake claims handling work for other insurers, it typically just received the initial notification of claim from the insured.  The appellant did not provide extensive underwriting support services to other insurers.  Instead it had an “insurer development team” which negotiated agreements with the other insurers and liaised with them more generally.  Historically the appellant has not received commission above the base commission from other panel insurers.  As noted, the timing of the payment of premiums was different. 

127.       Mr Pavey and Mr Godfrey both noted that the appellant had in the past provided more extensive risk pricing, underwriting analysis, claims handling and intermediary services to other insurers including to a Japanese insurer, Chiyoda a subsidiary of Toyota (which also owned the appellant until 2002) from 1997 to 2002 and Inter Hannover Limited from 2001 to 2006.

Overview of interaction of the appellant’s and Advantage’s business

Entering into insurance policies and underwriting function

128.       The appellant sold insurance policies in the UK on behalf of Advantage and other insurance panel members.  Customers could seek quotes for insurance cover through the appellant’s website, aggregator or price comparison websites (websites which provide for quotes from a number of insurers) or the appellant’s call centre.  Mr Pavey said that price comparison websites have increased in importance as a distribution channel; for the year ended 31 December 2013, around 89% of new insurance policies arranged by the appellant were sold through such websites. 

129.       The appellant had binding authority to conclude insurance contracts and renewals of contracts on behalf of Advantage, within the underwriting guidelines and limits set by Advantage.  Mr Charlton confirmed that, as long as the appellant was acting within the terms and conditions on which Advantage was prepared to allow it to accept risk, it was authorised to solicit and receive proposals and underwrite and accept business on Advantage’s behalf. 

130.       Mr Charlton emphasised, however, that Advantage had control of what risks or business could be underwritten/insured.  Advantage decided what risks to insure for what “risk price” or net premium.  Accordingly Advantage decided which types of driver, what types of car, what uses of the car in which locations and which types of disclosure by the insured were acceptable.  The “risk price” or net premium is the amount required by Advantage in order to (a) cover Advantage’s claims and other costs in relation to its insurance policies and (b) provide Advantage with a profit.  Mr Charlton and Mr Godfrey both said that it was entirely for Advantage to determine the net premium pricing, acceptance criteria and any changes although it did so using calculations and analysis partly produced by the appellant’s staff in Bexhill and partly by Advantage’s own staff in Gibraltar (as set out below). 

131.       Once the net premium was set it was inputted into an automated insurer pricing system, CDL.  Ms Johnson explained that it is typical for an insurer to have its own “rating tables” whether held on a pricing system provided by a third party, such as CDL, or operated by the insurer itself.  These comprise rules which are applied to the information provided by customers as fed into the system by the appellant’s staff.  If the customer is within the acceptance criteria set by the insurer, the system produces a net premium quote for the insurance cover.  Generation of the quote, therefore, indicates the insurer is willing to contract with that customer.  The initial net premium was then fed into another system, RTP, which is a flexible system enabling simple quick “real time” changes to be made.  More sophisticated changes could be made in CDL but they would take longer to implement.  The RTP system also contained the rules under which the gross premium was generated as set by the appellant. 

132.       Any changes to the net premium in CDL could only be made by Advantage (although acting through the underwriting support team at the appellant).  The retail team at the appellant, which set the gross premium, had no authority to change the net premium in CDL but could, as for other insurers, on the insurer’s instructions, make changes in RTP. 

133.       In the majority of cases the appellant entered into contracts with customers on behalf of Advantage on the basis of the quote generated by, and within the criteria used within, the automated system.  The brokers had to follow specific guidelines to refer particular customer risks falling outside of the set criteria to a manager at the appellant or Advantage.  Mr Charlton said it was very rare for referrals to be passed to the Advantage team in Gibraltar.  Mr Gumbrell said that in his experience there were typically less than five such referrals per month.

134.       Mr Gumbrell said that in cases falling outside the set criteria where a decision could be taken by the appellant, the appellant’s underwriting manager or director would decide how to manage the case by looking at multiple factors which were broader than those taken into account in the automated system, such as (a) any claims made by the customer (b) how long the customer had been insured and (c) how diligent the customer had been in relation to documentation and payment.  In such cases the premium and policy terms were calculated outside of the system using a spreadsheet. He said that this discretion was essentially confined to cases where an existing customer, who had been within the set criteria in the first place, had a change in circumstance, such as if he bought a new car over a specified value. 

135.       Mr Charlton said that the appellant was not otherwise at liberty to negotiate the risk price and acceptance criteria but merely accepted risk on behalf of Advantage.  Mr Charlton thought 95%, if not all, of the agency agreements between an insurer and an intermediary, were made on the basis that “the insurer has the right not to insure a risk if it feels it is not acceptable” and he said “that would be the same with Norwich Union, Royal Sun Alliance or whoever”.  He confirmed that the accepting of risk, on behalf of Advantage, took place in the UK.  He said that there was nothing unique in the appellant’s relationship with Advantage in that regard. 

136.       In his view it could be said that the appellant did not underwrite any risks.  All it did was, as he had explained, accept risks within the guidelines based on the questions that were put into the automated system (subject to the limited exceptions).  He described that as being the same as for every other broker or insurance intermediary in the personalised motor business - “it is all automated”.

137.       Mr Charlton did not accept that he was keeping a close eye on the appellant in accepting risks.  Accepting risks was what the system did.  As set out above, in effect, risk was accepted when the appellant’s operators/brokers entered into the contracts with customers in the UK. Mr Charlton confirmed that, at the point of sale of the policy/renewal to the customer, there was no intervention by Advantage (as Ms Johnson also said and was clear from the “script” in the bundles which the appellant’s call centre team used when dealing with customers).  Advantage monitored this only in the sense that its senior underwriter, Mr Powell, and another Advantage employee, Ms Faye Thomas, sampled the CDL/RTP system for Advantage to make sure that it was producing the correct net premium rate. 

138.       Ms Johnson said that the appellant’s sales systems did not prefer any individual insurer over another when providing net premium quotes to customers.  Mr Pavey said that the appellant provided premium quotations to price comparison websites based on the most competitive net premium quote received from insurers (whether Advantage or other panel members) but may quote a range of different policies each with different specifications and/or branding. 

139.       Mr Charlton noted that the call centre staff who sold policies did not have insurance training.  They were trained to be customer facing and to be able to use the system but they had no insurance awareness and did not see the insurance result of what they were doing.  It was their job to be “very friendly….to try and get the truth out…..to handle [customers] efficiently…. to try and give them the best price”.  As noted, under this essentially automated process, the system generated a number of quotes from Advantage and other insurers.  The operator recommended a quote to the customer largely based on price.  The operator may look at what the excess under the policy was, maybe “a few other bits and pieces”, but the decision was mainly made on price.  That is what most insurance is sold on. 

140.       Mr Charlton said it is not just Advantage which has the form of control he described; every other insurer has the same arrangement with every single intermediary in the country.  That is normal practice.  Advantage was simply the most competitive on the risks that Advantage quoted for.

141.       Mr Pavey said that the work undertaken by the appellant’s broking team in selling insurance policies and placing the customers on cover was generally the same for Advantage as for other panel insurers.  He said that, in each case, the procedure was essentially as set out above.  Ms Johnson noted that the appellant’s panel review committee assessed Advantage’s credit worthiness in the same way as for other panel members and, as noted below, changes to customer policy wording were negotiated as for all the insurers. 

Appellant’s human and technical resources

142.       The appellant had autonomy to decide the number of employees it required.  During the Period, the number of staff employed by it grew from around 700 to over 1500 as a result of the growth of its business.  The appellant had 962 customer facing staff in 2012 and 1,138 in 2013 including, as Mr Godfrey said, hundreds dealing with claims.  The appellant owned the website, it had the relationship with price comparison websites, it owned the call centres, the computer system which was used to provide quotes, the software system used for selling policies and for claims, the rating tables, anti-fraud software, customer database and the electronic and paper records of the insurance transactions and claims, which were physically kept in Bexhill.  Advantage had read-only access, during normal business hours, to the claims and policy software systems, which contained records of the customers and their claims.  It also had access to the appellant’s management information system which contained the extensive monthly reports prepared by the appellant.

Policy conditions

143.       The standard form insurance policy wording, and updates to it, were proposed by the appellant and approved by Mr Powell or referred to Mr Charlton.  Mr Charlton said that the majority of changes to the policy wording were driven by UK legal and regulatory requirements or to conform to changes in market practice.  Ms Johnson said that, in her experience, policy changes were negotiated with Advantage as with any other insurer. 

144.       Mr Gumbrell said that policy changes have become less frequent as the appellant has moved to a common policy wording for all its panel members.  He said that, however, as it was the majority panel insurer, Advantage’s interests had quite a lot of influence over the policy terms; proposed policy wording changes were sent to Advantage for review and approva1.

Setting the net premium – budgeting/forecasts

145.       Mr Charlton said that Advantage’s pricing of the net premium was driven by the budget set by the finance/actuarial teams in Advantage as approved by the board.  The Advantage board provided an overall profit target and the Advantage finance team and actuarial function modelled what price rises and premium volume would be necessary to achieve the target, given predictions about the market such as expected claims and premium inflation.  The process of setting the targets was originally run by the Advantage senior actuary, who built the forecasting model, although it involved discussions with the appellant and HIGL regarding the assumptions underpinning the analysis, in particular, as regards the estimated gross written premium targets.  When he was CEO, Mr Charlton had ongoing discussions with the board as the budget and business plan were developed, particularly through his weekly meetings with Mr Feetham. 

146.       Towards the end of the Period (from sometime in 2013 onwards) the appellant’s underwriting team had a greater role in producing the analysis to enable Advantage to set the targets.  Mr Gumbrell, who was directly involved in the process for setting targets in 2013 and 2014, explained that the profit target was based on Advantage’s loss ratio for the year (the cost of claims as a percentage of premiums) and gross written premium (the gross premium charged to the customer, before any deductions for premiums due to reinsurers and any commission).  He said that, at that time, the analysis carried out by the appellant for Advantage focussed on what net premium changes needed to be made to offset claims inflation having regard to expected future loss ratio improvements and changes in mix of customers. 

(1) As the first step the appellant’s actuarial team projected claims inflation based on historical trends and their actuarial models.  The appellant went through the analysis with Advantage for every type of claim (accidental damage, theft, personal injury and third party property damage) looking at the frequency and severity of claims.  This was debated and challenged in a series of meetings with the Advantage managing director and relevant members of the appellant’s team. 

(2) To determine the gross written premium target the appellant analysed the net premium rate changes that Advantage budgeted for against what it thought the market would do based on market data on market loss ratios and the claims inflation predicted by the appellant’s actuarial team.  These assumptions were fed into a model produced by the appellant’s finance team, with the appellant’s commercial team’s planned initiatives and commission changes. The output was estimates of movements in Advantage’s policy because of its pricing relative to the market.  Account was taken of systematic changes planned by the appellant such as changes to the channels through which it distributed policies.

(3) This culminated in the production of the targets and accompanying planned rate changes for approval by the Advantage board.  The whole process typically took about a month.

147.       Mr Godfrey noted that, when he was managing director of Advantage, the business plan was set in the budgeting process for the next financial year.  It appears that from 2014 there was a more specific rate change plan which set out the expected increase to be made in the net premium from a specified time.  Advantage and the appellant’s underwriting support team would work on the basis that the planned rate changes would be implemented (assuming there were no market developments which may require a further change).  He described the process in similar terms to Mr Gumbrell noting that the forecasting model used by the appellant was originally built by the Advantage senior actuary.  He also noted that the appellant also inputted into the plan projections based on its gross premium (and other product) pricing plans.

Setting the net premium

148.       Mr Charlton said that in setting the net premium pricing for Advantage, its senior actuary, Mr Morales, used actuarial techniques, using a base rate (a fixed premium of a set amount) with various multipliers which adjusted the premium by reference to each particular risk.  Advantage used a large number of risk relativities (around 50), being factors such as the age of driver, the age of car and the number of accidents.  Advantage monitored each risk by looking at performance metrics that were broken down by these risk relativities or combinations of them.  Mr Charlton said that the results were presented to the rating analysis committee (“RAC”) and he or Mr Godfrey made the decision as to what rates were to be charged to achieve the business plan. 

Subsequent changes in net premium

149.       As regards subsequent changes in rates, from February 2009 to September 2010, Mr Charlton received underwriting analysis and rate change proposals directly by email from Mr Godfrey as the appellant’s underwriting director.  This was similar to the information later provided at rate presentation or RAC meetings.  Mr Charlton typically gave a rough guide of the rate change he was expecting.  He communicated rate changes direct to Mr Godfrey.  The procedure changed as the volume and complexity of business underwritten increased and it became necessary to have the appellant’s team present their own analysis rather than provide it via Mr Godfrey. 

150.       From around September 2010 onwards, each month the appellant’s staff presented a report to the RAC or rate presentation meeting analysing Advantage’s underwriting performance.  Any proposals for changes in rates were suggested by the appellant’s team also for consideration at the meeting but, as before, Mr Charlton suggested a guideline of the price Advantage were looking for.  He and the actuary looked at the loss ratio for the previous month and using pricing strength models estimated a final loss ratio (the ratio once the claims had fully matured).  The model was originally developed by Mr Kelsey and Mr Morales developed a more sophisticated one.  Before that Mr Charlton used his own experience to estimate the likely development of the loss ratio.  Depending on where the ratio sat relative to the budget Mr Charlton would know whether or not Advantage needed to make a net premium rate increase.  All such decisions were made by him or Mr Godfrey (once he was the managing director) in Gibraltar in line with the business plan requirements set by the Advantage board. 

151.       Advantage staff had full access to the reports and data analysis undertaken by the appellant if they wished to verify an item.  Mr Charlton confirmed that the appellant had no discretion to alter the net premium without instruction from Advantage.  Advantage’s underwriting team monitored the outputs of the pricing system to ensure the correct net premium was being used.  If an error was identified it was discussed with the appellant.

Appellant’s budget and group budget

152.       The appellant had its own annual targets and budget which was approved by its board.  Mr Pavey said that the annual budget process for the appellant was led by the finance team but involved interaction with all the appellant’s business teams.  The income targets included expectations of new business and renewal volumes across the product range taking into account assumptions on market conditions.  Assumptions were used to consider what broking income per policy would be achieved, including the base commission to be added to net premiums.  The cost base was budgeted at a detailed level which involved each sector within the appellant setting targets.  The budget process was replicated in developing a three-year plan, albeit at a higher level and with a broader set of assumptions.

153.       As part of the process, discussions were carried out between the underwriting teams at Advantage and the appellant and with the appellant’s broking team.  The broking/retail team (headed for part of the period by Ms Johnson) provided information as to how the market might evolve over the following year and modelled predictions of the volume of sales of policies.  Mr Pavey noted that this analysis was also carried out in relation to the other panel insurers.

154.       He confirmed that, as Mr Godfrey said, the appellant’s finance team provided Advantage’s finance team with an estimate of the amount of business that the appellant expected to write the following year.  The figures and forecasts produced by the appellant were assessed by Advantage and used in the formulation of its business plan.

155.       Since April 2012, once Advantage’s business plan had been determined in Gibraltar, it was consolidated with the appellant’s business plan at a group level.  Senior executives within HIGL then discussed the group budget with both the appellant and Advantage.  Prior to that, as Advantage was not part of the same group of companies as the appellant, there was no such consolidation.

Setting of commission

156.       Mr Charlton said that (as Mr Pavey and Ms Johnson also explained) the appellant added its commission to or, very occasionally, subsidised the net premium price depending on what it felt would optimise the return for its business.  That adjusted amount, plus IPT, was the gross premium amount charged to the client but Advantage always received the net premium which it had set.  It was entirely Advantage’s job to produce that net premium.  The appellant had no authority or power to change that price.

157.       Mr Charlton noted that the appellant, as the retail side of the business, was also concerned with “add ons”, such as whether to sell personal accident cover or to offer other benefits such as car breakdown insurance.  That had nothing to do with Advantage.  The insurance part of the product was signed off and agreed with Advantage.  As noted, Advantage signed off on the policy wording.

158.       Gross premiums were set by the appellant without requiring authorisation from Advantage or any other panel insurer although the appellant sometimes notified insurers of major changes.  Ms Johnson said that the team’s aim, in setting the commission, was to “maximise value for the appellant from the net premiums set by the insurers” through achieving “the right volume with the right value”. 

159.       Mr Pavey said that, as for all insurers it acted for, the appellant acted as a broker, which used its own “propensity models” to consider whether a customer was likely to take other products such as premium finance, breakdown or legal cover and whether a customer was likely to renew the following year.  Using these propensity models, the appellant set a commission rate that optimised the “customer lifetime value” for both the appellant and Advantage (although before around July 2011 the appellant’s systems only considered that value for the appellant).

Management of customer relations/complaints

160.       The underlying management of the customer relationship was for the appellant although serious complaints were referred to Mr Charlton for him to deal with. The rest of the complaints were handled by the appellant within the guidelines set by Advantage.  Mr Charlton said he probably saw half a dozen complaints a year, at any rate, a very small number, and they would tend to be very serious complaints. Complaints in relation to claims handling were also dealt with by the appellant.

161.       It was drawn to Mr Charlton’s attention that the policy document said that the customer should first raise complaints with the appellant and, if the appellant could not deal with the complaint and the customer was still dissatisfied, the customer could write to the insurer.  Mr Charlton confirmed that customers could contact Advantage but the complaint would be referred back to the appellant to deal with unless it was one of the very serious ones he had referred to.

Mr Charlton’s evidence on interaction of the businesses

162.       It was put to Mr Charlton that Advantage’s ability to determine the conditions under which the appellant could underwrite insurance business and the net premium rates effectively allowed Advantage to determine what business the appellant could undertake.  Mr Charlton said that, whilst Advantage determined what business the appellant could place with Advantage, the appellant could in fact place a piece of business with any other insurer acceptable to it; as a broker it could use any other insurance company on the panel.  It mainly used Advantage but it had a choice of who it placed business with.  It just so happened that Advantage was the most competitive insurer in a number of situations.

163.       Mr Charlton continued that Advantage was not stopping the appellant using any other insurer and going into any other product areas.  He noted that the appellant dealt in home insurance which Advantage did not underwrite.  It could do what it needed to do for its own business requirements as decided by its board. 

164.       Mr Charlton noted that, as all insurers:

“they were always looking for business which they feel they can price, because they understand the risk attached with it, what liabilities would come with it and can work out a price.  If you cannot work out what the risk and the price is you cannot underwrite it…..there is nothing unique in the Hastings/Advantage relationship in that area.”

165.       Mr Charlton said that the use of the appellant as intermediary was “a best use of capital question”.  He decided that the business with other intermediaries was “not as good as the business with Hastings for Advantage and to use the capital, scarce capital resources at that time, not being a public company ……just owned by individuals to use that capital to write business via Hastings as opposed to other intermediaries.” 

166.       It was put to Mr Charlton that, whilst the risk side of the business was very important, equally, without the retail element, the distribution channels and sourcing and retaining of customers, there would no insurance business either.  Mr Charlton said that the fact the appellant carried on the retail element did not mean Advantage was not carrying on the insurance business.  He noted that the appellant was the distribution channel.  He thought that, if you looked at the market place, there was not a single Lloyds of London business that actually had direct access to “on-the-street markets”. He could not think of one.  But that did not mean they were not insurance businesses.  He noted that things have changed over the years.  When he first went into insurance every single insurance company dealt via intermediaries or brokers.  A number of insurers now deal direct.  At the time of the management buy-out in February 2009, he would have much preferred to turn Advantage into a direct business but that did not happen and he noted that the appellant still has the opportunity to broke business to other insurers.  He said that an insurance company which only deals with the public via intermediaries is no less of an insurance business.

167.       It was put to him that, as Advantage’s only intermediary was the appellant, Advantage would not have an insurance business without the appellant.  Mr Charlton said that without the appellant, Advantage would simply have to find a different distribution channel but that did not affect Advantage’s insurance business.  He agreed that, at this point in time, one could not operate without the other.  But either of them could obtain what the other one provides from somewhere else.  The appellant could build up a relationship with another insurer.  Advantage could build up a relationship with another distribution channel, or enter into direct business itself. 

168.       It was put to him that whilst Advantage could buy in certain services such as a call centre and claims-handling function, it would be difficult to replace the underwriting support services provided by the appellant.  Mr Charlton noted that these were basically data and statistical services which could be bought in from elsewhere.  He noted that “various companies try to sell…those services, particularly actuarial companies”.  He said:

“the system produces a wealth of data and information.  And there are companies out there more than willing and happy to be paid to sort that data and information out for you…….I am happy with the way we do it now.  I think it gives us some USPs [unique selling points] why..….going back to 2009, we are far more successful than we ever thought we would be.  And part of that is the business model we’ve got and the USP it gives us.  And that is Advantage….standing separately to Hastings, both of them doing their best …….for themselves, which then builds a better business for the group.”

Function of claims teams

169.       As set out in the services agreement and related claims handling guidelines, the appellant had authority to deal with small claims provided it acted within the parameters set out in the guidelines.  Large loss claims could only be dealt with subject to instruction from Advantage and were dealt with at the LLC.  The Advantage claims director recommended the threshold for large loss claims for approval by the Advantage board. 

170.       Mr Charlton described the main role of the Advantage claims team as being to manage the handling of large loss claims, to monitor claims payments and to manage the day-to-day relationship with the reinsurers.  The team carried out audit checking of the appellant’s work, in particular, to make sure that claims were reported and settled properly and that claims leakage was measured and controlled. Mr Lee described claims leakage as arising where unnecessary payments were made in respect of claims either because the claim was not settled in accordance with best practice or the terms of the policy, or subjectively, where a lower settlement might have been achieved.

171.       Mr Charlton or Mr Godfrey, Mr Eagar, Ms McKeever and the senior actuary at Advantage typically formed the LLC. The meetings were usually also attended by various members of the appellant’s large loss team.  These and the other meetings described below were held by a video conference between the appellant’s offices in Bexhill and Advantage’s offices in Gibraltar.  One of the attendees of the LLC meeting had to be either the Advantage managing director or the Advantage claims director.  Although senior employees of the appellant typically attended the LLC meetings, there were occasions when no representative of the appellant attended (as shown for example in the minutes of the meetings in June 2012 and in November 2013).

172.       The LLC set the level of the reserves for large loss claims and approved the required actions, normally after a discussion and consensus had formed between the LLC and the appellant’s attendees (who made recommendations).  The reserve set by Advantage sometimes differed from the appellant’s proposal.  Also the LLC decided what communications needed to be made to the reinsurers (pursuant to Advantage’s reporting obligations under its reinsurance contracts).  

Underwriting support services

173.       Mr Charlton said that the appellant’s other main function was to provide analysis through the underwriting team.  He described the team as “statisticians …who produce lots of different statistical information.  It would be one, two, three way analysis.  So it would look at… females, times, age, living in so-called post codes.  One of the factors we use a lot in rating is geo demographic rating, social economic groups.” This required the appellant to provide Advantage with a whole host of standard reports on a regular basis.

174.       When Mr Charlton was first at Advantage he could go into the appellant’s systems and see what business was written on that day but that was impractical and not much use.  What the insurer was really concerned with was the experience and the output of the insurance over time.  If “you did things too quickly, if I changed a factor today and I looked at it tomorrow, I have had no earned exposure at that new factor.  It would take at least six months to get a reasonable view of any change in claims frequency, it would take at least a year, maybe longer, to get an idea what the average claim change would be for that mix. So it is a long-term thing”. The intellectual property of that analysis sat in Advantage and both Advantage and the appellant, equally, “would do ad hoc type analysis, more in depth analysis of an area to things that are beyond the regular reports”.

175.       When Mr Godfrey was the appellant’s underwriting director, he oversaw the underwriting analysis work carried out by the appellant, which he said involved the following:

(1)  account performance monitoring, which involved providing Advantage with weekly management information about the loss ratios, sales, customer premiums, and claims for categories of business;

(2) running automated models, performing calculations, and providing data and recommendations for net premium rate changes for Advantage; and

(3) investigation and analysis to support product development and exploration of different strategic areas and directions in relation to insurance policies underwritten by Advantage, such as exploring the profitability for different customer segments not within Advantage’s core customer book.  Mr Godfrey sometimes required specific analysis to be undertaken as discussed below.

176.       Mr Charlton had oversight over Mr Godfrey’s work whilst he was the underwriting director at the appellant but he reported directly to Mr Lee (and before his appointment, his predecessor).  Mr Charlton’s oversight extended to satisfying himself that the underwriting support services performed by the appellant for Advantage were being carried out properly. 

177.       As Mr Charlton explained above, from September 2010 proposals for rate changes were made and discussed at monthly rate presentation and RAC meetings.  Changes to risk acceptance criteria were also considered at these meetings.  Mr Godfrey said the process for considering changes was the same whichever type of meeting changes were considered at.  The meeting attendees were typically: from the appellant, Mr Godfrey (when he was the appellant’s underwriting director), Mr Lee, the chief actuary at the appellant and the team analyst responsible for working-up any underwriting change proposals and, from Advantage, Mr Charlton or later Mr Godfrey, Mr Pablo Morales (the senior actuary) and the finance director or a member of the Advantage underwriting team. 

178.       Advantage decided on all changes to net premium and risk acceptance criteria as set out in further detail below.  As set out above, the Advantage underwriting team made sure that the rates that were entered by the appellant on the pricing systems were the rates provided by Advantage, that they were properly calculated on the system and were consistent with the expected result of the rate change.  They also checked that the correct policy rates were given to customers at the quote stage and that the appellant only accepted insurance business that was within the parameters of Advantage’s underwriting guidelines.  These were checked by the senior underwriter using sampling and system generated checks.  In carrying out the validation work, Advantage used the appellant’s underwriting IT systems and rate models for the commercial vehicle and motorcycle books built by the Advantage team.

179.       Advantage’s underwriting team had some interaction with the appellant’s underwriting support team in that, if an issue was discovered (such as incorrect entry of rates on the CDL system), this was reported by Mr Powell to the appellant for it to investigate.

ORM

180.       The ORM met on a monthly basis, as Mr Charlton described, to review the service provided by the appellant’s claims teams, monitor their performance, and assess whether enough staff were employed and were carrying out the required functions effectively.  Mr Eagar also reported claims handling matters (such as significant large loss claims and the appellant’s operational performance) to the Advantage board.

181.       The main attendees at the ORM meetings were the senior staff of Advantage along with Mr Lee and his senior staff.  Before Mr Lee’s appointment the main attendees from the appellant were the insurer services director, Mr Godfrey, as the appellant’s underwriting director at that time, the underwriting operations manager and the claims director.

182.       Between September and November 2010, Advantage started to hold monthly “Insight” meetings at which the appellant’s anti-fraud work was reviewed.

Advantage’s finance and budget

183.       Since the management buy-out in 2009 and up to and including the final budget in which Mr Charlton was involved in 2013, all the Advantage budget work was done by Advantage in Gibraltar as set out above.  As noted the appellant became more involved in the process for setting targets at the end of the Period in 2013. 

184.       Mr Godfrey explained that during his time at Advantage, Advantage’s finance division also prepared the budgets and forecasts which were used for the group-level work done by the group finance team.  It prepared monthly profit and loss reports which were included in the Advantage board packs.  The data for these reports came from the appellant’s system.  Advantage queried movements in the loss ratios with the appellant at the RAC meetings.  The profit and loss reports were used as the basis for some of the strategic decisions made by Advantage such as net premium rate changes.  Advantage determined the loss ratios used in the profit and loss reports based on the quarterly reserve review of ultimate loss ratios and interim management information for months between the quarterly reviews.

Advantage’s actuarial team

185.       During the Period, all the actuarial reserving work, that is determining what Advantage’s claims reserves should be for the purposes of its accounts and regulatory filings, was carried out by Advantage’s own actuarial resources in Gibraltar and independently by an external actuary.  Advantage was required to report its reserves in order to comply with the FSC.  The reserves were also monitored by its auditors, KPMG.  The Advantage board reviewed the reserving work carried out by or on behalf of Advantage.  The process has changed since Mr Charlton retired from his role as CEO; the appellant then provided some reserving services to Advantage.

186.       Mr Charlton noted that in 2009 Advantage employed Mr Kelsey as an actuarial consultant while they were looking for an actuary to join the business.  He noted that it became pretty obvious that fraud was becoming a concern and there was a need and a desire by Advantage for the appellant to look more into the fraud element when they were taking business on and when they were paying claims.  Since that started it has become a fairly big area for the appellant’s claims department.

187.       Advantage has had an employee employed on a full-time basis as actuary located in Gibraltar since 2011.  Mr Godfrey said this provided Advantage with a better understanding of its liabilities through the quarterly internal and half yearly external reserve reviews.  The actuary also did capital/solvency calculations, oversaw regulatory capital work on the implementation of EU Solvency II pillar 1 requirements and reviewed certain technical aspects of rating changes and loss ratio movements.

188.       From 2011 to 31 December 2014, Mr Morales was Advantage’s senior actuary.  During the Period, reserving reviews were carried out by Mr Morales.  He had regular meetings with the appellant, including a weekly call with its actuarial pricing team to discuss what that team were working on as part of the underwriting support services.  He also visited the appellant’s offices in Bexhill four or five times a year, for example, to examine particular pricing projects (such as new or updated claims cost and pricing models and special projects such as the implementation of changes arising as a result of the EU Gender Directive or a vehicle group review) or to speak to the appellant’s claims, underwriting and pricing teams when carrying out quarterly reserves reviews.

189.       In 2014 Advantage set up a reserving committee which recommends the levels of reserves for review by Advantage’s audit and risk committees and review and approval by the Advantage board.  Mr Charlton is not a member of the committee but attends its meetings. The appointment of the external actuaries is an Advantage board decision. 

190.       The audit and risk committees of Advantage both comprised Mr Charlton, Mr Feetham and Mr Pierre Lefèvre (a non-executive director of Advantage) and each met four or five times each year.  These committees reviewed the audited accounts, including consideration of the recommended reserves and business and operational risks. The recommendations of these committees were presented to the board of directors of Advantage.  Advantage’s CEO reviewed the appellant’s account management information regularly in order to see whether any issues needed to be raised with the support team at the appellant that they may not already have picked up.

Investment

191.       All investment work was carried out by Advantage’s employees in Gibraltar and investment decisions were taken and all related contracts were approved and signed by the board of Advantage.  The investment portfolio was professionally managed, from early 2009, by IAG Asset Management and Union Bancaire Privée and, from 1 July 2011, under agreements with Aberdeen Fund Management Limited (“AF”) and Deutsche Asset Management (UK) Limited.  Under the agreement with AF only Advantage staff were able to use AF’s electronic reporting platform and the only signatories authorised to act on behalf of Advantage were the directors of Advantage.  At all times, the investment portfolios managed on behalf of Advantage were required to be admissible under the supplementary regulations to the Gibraltar Insurance Companies Act 1987, as adopted by the FSC.

192.       In addition to portfolios which were managed on its behalf, Advantage also made direct investments in a number of fund vehicles and property ventures. Advantage received investment management advice from P-Solve Meridian on its overall investment strategy.

193.       All cash (other than investment monies which were not used for transactions) belonging to Advantage was held in current accounts provided by banks whose branches are in Gibraltar. The directors of Advantage were the only persons authorised to give instructions regarding payments from these bank accounts.  All transactions and payments between the appellant and Advantage were made from or to (as applicable) a bank account of Advantage held in a Gibraltar branch of a bank. 

Reinsurance

194.       Advantage had two reinsurance programmes:  excess of loss reinsurance, pursuant to which reinsurers met claims in excess of £0.5 million (now £1 million)); and quota share reinsurance, pursuant to which reinsurers were liable for 50% of Advantage’s claims.  Advantage handled claims notifications and paid the reinsurance premiums to its reinsurers.  The panel of quota share reinsurers used by Advantage expanded to ten by 2013.  Any proposed change in terms of quota share or excess of loss retention had to be approved by the Advantage board.

195.       Mr Charlton and later Mr Godfrey oversaw Advantage’s co-insurance and reinsurance arrangements and negotiated the reinsurance contracts.  Since he joined the appellant in June 2011, Mr Lee has assisted with these discussions because, from previous employment, he had strong personal contacts with Advantage’s reinsurance broker and with reinsurers.  Mr Lee said that he kept Mr Godfrey informed of developments through the process, as Mr Godfrey confirmed in his evidence.  Mr Lee noted that he sometimes provided a verbal reinsurance update to Advantage’s board.  He sought reinsurance arrangements that complied with the parameters set by Advantage’s board.  Both he and Mr Godfrey confirmed that he did not have authority to enter into any contracts on behalf of Advantage.

196.       Mr Godfrey described that, in the third quarter of each year, Advantage completed a questionnaire and supporting documents for the reinsurers which formed the basis of negotiations.  The questionnaire included account details broken down into various factors (such as age and no claims bonus for previous years and those projected for the next year), projected gross written premiums and the rate movement for the next year.  The content provided included analysis done by the appellant on issues such as postcode exposure and large loss development. 

197.       Mr Godfrey usually presented a reinsurance proposal to the Advantage board ahead of the year end renewal.  The board normally gave authority for him to enter into reinsurance contracts within certain parameters based on recommendations from the non-executive directors who, particularly Mr Charlton and Mr Lefèvre, had extensive reinsurance knowledge.  They related to (1) the structure of the reinsurance contract (for example, that the retention of loss (the level of loss which must be borne by Advantage before the loss is borne by the reinsurer) should not exceed a specified amount) and (2) the reinsurance premium pricing increases that Advantage was working to in its budget forecast. If the terms were outside of the parameters, he would need further board approval.  The renewal proposal received from the reinsurer was presented to the board and, if it was outside the parameters, further board authorisation would be required.  All of the reinsurance contracts were signed by Advantage in Gibraltar. 

198.       Mr Charlton noted that he occasionally met Advantage’s reinsurers in London. The reinsurers sometimes visited the appellant’ offices in Bexhill to look at specific claim files (as set out below).  Mr Charlton said that in the insurance industry, visits by reinsurers to the insurer’s outsource provider are part of the normal process to understand fully the losses for which they may be liable.  

Regulatory and other

199.       Advantage was supported in carrying out its regulatory obligations by HIGL’s company secretary and his team which assisted with taking minutes of committee meetings, provided secretarial support and information about the group and relevant UK legal or regulatory developments.  Mr Godfrey dealt with the FSC in Gibraltar directly regarding EU Solvency II preparations, regulatory filings and requests for information.

200.       Advantage was responsible for filing its regulatory returns, including its annual insurance returns, to the FSC on a quarterly basis.  These returns set out its financial and solvency position.  Advantage filed Annual Insurance Returns with the FSC which were completed on the basis that it was providing motor insurance on an “EEA Service Basis”.  Advantage paid all levies and other similar payments (as provided for in the agreement) and paid IPT due to HMRC on its insurance contracts from its NatWest current account in Gibraltar. 

201.       The appellant did not have any direct contact with the FSC in Gibraltar.  However it assisted Advantage in complying with its regulatory requirements by carrying out work in relation to regulatory requirements affecting Advantage under a programme set up to deliver EU Solvency II compliance for Advantage which at the time of the hearing was on-going.  Mr Godfrey said that all key decisions on this project would be made by the board of Advantage who regularly considered the progress of the project

Role of Mr Charlton

202.       Mr Charlton’s main responsibility was running Advantage and managing the underwriting side of the business closely, given that was his background.  His other responsibilities included ensuring that regulatory returns were filed with the FSC in Gibraltar and that budgeting and forecasting was done, preparing board packs for Advantage and attending board meetings.  Every day varied but typically he had various meetings, some with the appellant, using video conferencing, mainly with personnel in the office at Bexhill.  He very rarely visited the Bexhill office and did not do so at all in 2012 or 2013.  Occasionally he went to London for meetings with reinsurers or the external actuaries.  He also had meetings with the other major shareholders of HIGL on shareholder related business.

203.       From February 2009 Mr Charlton rarely spent time in London, and he estimated he spent less than 1% of his time working on Advantage business in the UK.  Around 95% of his time was spent on Advantage business in Gibraltar, with HIGL shareholder or board meetings occupying the remainder. No board meetings or decision making for Advantage took place in the UK, as this was all conducted in Gibraltar.

204.       Mr Charlton typically engaged with the managing director, finance director, claims director and the senior actuary on a daily basis and with the rest of the staff on a regular basis.  He would usually see the chairman of Advantage, Mr Feetham, on a weekly basis.

205.       When Gary Hoffman was appointed as CEO of HIGL on 1 November 2012, Mr Charlton usually had a weekly call or video conference with the appellant involving Mr Hoffman.  Mr Charlton described this as the main management meeting whereby typically four to six executives of the appellant discussed their progress against their business plans and reported at a high level regarding their areas of responsibility. 

206.       Mr Charlton also attended the weekly LLC meetings and the monthly ORM, and RAC meetings, the reserving committee and the risk and audit committees.

207.       Mr Charlton also interacted with the reinsurers and the external actuary.  On a less regular basis he liaised with the investment managers and occasionally, along with the finance director, Advantage’s auditors, KPMG.

Board and executive meetings

208.       From the evidence set out below, we accept that the appellant’s and Advantage’s boards operated independently of each other, albeit that there was some interaction in terms of the provision of information and updates by employees of one of the entities to the board of the other.

Board meetings of Advantage

209.       Advantage’s board meetings during the Period were held in Gibraltar.  When Mr Godfrey was at Advantage, the board comprised, as executive directors, the finance director, Mr Eagar and Mr Godfrey and, as non-executive directors, Mr Feetham, Mr Charlton and Mr Lefèvre (who in the past held various senior executive roles at the group). 

210.       The board discussed any strategic decisions made by Advantage including investment decisions, reinsurance arrangements, finance and budget policies, outsourcing and insurance business, gross written premium volume, profit position, forecasting, reserving, claims, risk, internal audit and compliance.  The senior staff in Gibraltar had power to take decisions that did not require board approval, such as setting large loss claim amounts.

211.       Mr Charlton noted that, when he was CEO, he reported Advantage’s underwriting performance to the Advantage board.  The board delegated the review of Advantage’s business plans to Advantage’s executive management.  Advantage’s actuary decided which areas Advantage and the appellant should focus on, for example, by producing analyses that highlighted poor performing areas of the book.  In particular, Mr Charlton recalled a specific instance when he highlighted that older vehicles in lower social/economic segments had a particularly high loss ratio.

212.       Some of the appellant’s directors sometimes attended Advantage’s board meetings.  Mr Charlton said this was an occasional occurrence.  Mr Godfrey described their attendance and that of Mr Hoffman, the CEO of HIGL, as “regular” once the appellant and Advantage came under the ownership of a common corporate parent in 2012.  They both said, however, that their input was limited to providing information and assistance on matters, such as the appellant’s service performance (such as staffing levels in its claims or underwriting support teams) or, specific items such as performance results, the profit and loss of Advantage or the appellant and any issues relating to the appellant that the board should be aware of as its outsource provider.  

213.       They confirmed that the appellant’s directors did not participate in any decision-making by the board of Advantage.  All decisions were taken by the directors of Advantage only during the meeting, following discussion amongst themselves. 

214.       Mr Godfrey attended these board meetings when he was the underwriting director at the appellant to give the Advantage board an update on the latest account performance figures.  He described this as just a general underwriting update on what the appellant was considering.  He said that “obviously at the board meeting we would not agree a rate change as such, but I would just update them on what work is being done in the underwriting team in Hastings and get a steer from the board, you know, if there is anything else they want us to look at or we should be thinking about”.

215.        Mr Lee confirmed that he attended Advantage’s board meetings, either physically or by phone, as a representative of the appellant. He also said that only Advantage directors participated in decisions made by the board.  Generally, Advantage’s non-executives questioned him (as the executives had opportunities to ask their questions at the monthly ORM) on claims patterns and trends in the insurance market that he and his teams were seeing.  He may also be asked to report on significant matters relating to the areas of the appellant’s services for which he was responsible such as audits on the claims department. 

216.       He noted that if, when questioned about an issue or trend that Advantage had picked up in the management information, he was unable to explain why it was happening (for example, whether it occurred because of a process change in service delivery, a general change in the insurance market or a specific action by another market participant), he would be given an action point to investigate further. If the appellant was unable to explain an issue, Advantage may consider changing the net premium rates or adjusting its exposure to the market to mitigate against the risk.

217.       He noted that, if the issue arose from his/his teams’ poor performance, he would come under pressure at the appellant’s board meetings and from the HIGL executive committee and at the ORM.  He said that although he thought it very unlikely, Advantage could ultimately decide to use a different outsource company if the appellant’s performance remained poor.

Board meetings of the appellant

218.       Mr Pavey gave evidence that the appellant held a board meeting approximately every two months during his time on the board.  Advantage directors were not present at any of these meetings, other than when Mr Charlton attended meetings of 30 July 2009 and 10 February 2011.  At the later meeting Mr Charlton participated only in a discussion concerning the EU Gender Directive by videoconference; he only attended for discussion of that single item on the agenda.  The minutes of the meeting of 30 July 2009 noted that that if any matter was discussed “for which it was inappropriate for Mr Keith Charlton to take part then he would leave the meeting until such discussion had been concluded”.  Mr Pavey said that Mr Charlton was not involved in the decision making of the appellant’s board on either occasion.

Executive committee

219.       The group held executive committee meetings which, as Mr Lee said, were a forum for the various executives to update Mr Hoffman, as the group’s CEO, and each other with developments in the various business areas and for strategic group-level decisions to be discussed.  Mr Lee said that, in his experience, the meetings were attended by the CFO of HIGL, HIGL’s company secretary, typically five members of the appellant’s executive team, Mr Godfrey from Advantage by telephone or videolink and in the past, Mr Charlton.  Matters discussed included market prices and what the appellant intended to do on commission, the grade of service on the telephones and recruitment issues.  Advantage reported on matters such as claims inflation and what it intended to do with net prices.  Mr Lee attended these meetings either on the phone or in the offices at Bexhill. 

220.       Mr Lee noted that some matters relating to the appellant, such as major changes in its commission, were subject to approval at the executive committee.  Ms Johnson said that changes above about 5% had to be referred to the committee but, in her experience, there had never been such a referral. 

221.       Some matters relating to the services the appellant provided to Advantage were discussed because they could have implications for the loss ratio in Advantage.  Mr Godfrey may be asked for his views and, if he objected, the matter would either not go ahead or be referred to the appellant’s and Advantage’s boards of directors.  The relevant executive at the appellant responsible for that matter would deal with its implementation.  As Mr Godfrey and Mr Charlton also said, however, the executive committee did not take decisions on behalf of Advantage at these meetings. Mr Charlton said that on occasions he was asked for his views and gave high level updates as regards Advantage, but Mr Hoffman and the other attendees from the appellant made the decisions.  Mr Godfrey confirmed that essentially these meeting continued in the same way when he was managing director of Advantage. 

222.       Mr Lee confirmed that the committee had no control over the setting of net premiums.  Changes in commission were discussed but he did not ever recall the committee actually ever telling the retail heads what commissions to charge.  It was much more a case of each of the business entities updating each other and discussing the influence of each other’s proposed changes, but not actually enforcing decisions.

223.       So the executive team reviewed the budgets for each of the businesses, as each one needed to know what the other was doing.  He thought Advantage primarily focussed on how much business it thought it could profitably write in the market conditions.  The appellant would see that as how much business they were going to get from the net rates of Advantage.  He said that, to put that into context, if that met their growth ambitions, for example, that was fine.  If it was not enough, then they would potentially look to add additional underwriters to the panel, sell different ancillaries, change the products in some way, or change their marketing.  But they would look at what their primary underwriter was going to do and then decide how they were going to adapt to that.

224.       Mr Lee said that the executive committee did not impose a budget on either the appellant or Advantage.  That was a decision for the individual boards.  He noted that some papers that were intended to go to the various boards were presented first at the executive committee and the committee was used as a forum for the businesses challenging each other, before the papers went to the individual boards for discussions and agreement or not.

Account monitoring and changes to the policy, the net premium and risk criteria

225.       Mr Gumbrell explained that Advantage had three main levers that it could use to achieve its loss ratio and gross written premium targets:  it could change net premium rates, it could amend the acceptance criteria for customers and/or it could make changes to the policies (for example, to the exclusions, level of cover or excesses).

226.       Mr Gumbrell said that changes to the policy wording were probably made a couple of times a year.  The team at the appellant liaised with Advantage by email or telephone about those changes.  The net premium rate and risk acceptance criteria changes could be changed as often as weekly, although historically the frequency of rate changes had tended to be monthly.  Mr Godfrey said that the proposals were to maintain Advantage’s performance in the market and, when there was a rate change plan, to implement the planned rate changes. 

227.        The procedure for changes made until September 2010 is set out above.  The below largely relates to the position from September 2010 when proposals for such changes were presented by the appellant at the RAC or rate change presentation meetings (the process in either case being the same).  At all times in the Period decisions on whether to make such changes were made by Mr Charlton or, when he became managing director of Advantage, Mr Godfrey, in line with the business plan requirements set by the Advantage board.

Changes to net premium and acceptance criteria – account monitoring

228.       The first part of the process was for the appellant to track Advantage’s performance against its targets for profits and growth by monitoring loss ratios and the gross written premiums.  As Mr Gumbrell set out, the appellant’s team prepared a weekly set of reports, which tracked matters such as customer mix and the number of policies underwritten, and a monthly set, which tracked matters which did not need to be monitored as frequently, such as the number of claims.  Advantage received all of the main reports as and when they were prepared.  The reports were collated with other management information and included in the pack that was provided to Advantage in advance of RAC meetings. 

229.       The appellant’s team analysed the reports to identify potential opportunities and risks for Advantage.  They tracked the mix of customer business and looked for inconsistencies between the cost of claims and the net premium.  Advantage, mainly through Mr Godfrey and Mr Charlton, also made requests for analysis in particular areas. 

230.       For example, the analysis may reveal that the loss ratio in a particular customer segment of the book (such as a particular category of policyholders within a specified age group and area by postcode) was particularly low, such that Advantage may be able to reduce their net premium rate for that segment.  On the other hand, if the information indicated that Advantage’s volume in a particular segment was increasing unexpectedly that may indicate that Advantage had mispriced the risk.  The team also looked for opportunities for Advantage to expand its risk acceptance criteria in areas with limited historic claims through “test and learn” activity.  They tracked this business very closely to ensure Advantage could achieve an appropriate mix of customers and a level of net premium so as not to compromise its target loss ratio.

231.       Mr Gumbrell’s team also included in the reports information from the appellant’s retail team as regards the impact of the appellant’s commission charge on volumes and Advantage’s mix of business.  This had to be monitored to ensure that Advantage did not breach the terms of its reinsurance contracts as to the mix of business it could insure. 

232.       The reports were then considered at the monthly RAC meetings at which Advantage reviewed the performance of the various accounts within Advantage’s book of business (private car, motorcycle, commercial vehicle and telematics). That involved looking at the gross written premium against the budget and then, in greater detail, the volumes of new customers, cancellations, new claims and the types of claims and how these various measures were tracking against the more detailed targets for Advantage’s accounts.

233.       Mr Gumbrell said that Advantage sometimes asked his team to conduct further analysis and to come back with proposals to fix a trend that Advantage may have picked up in the management information, such as, an increasing rate of cancellations within a particular part of the book or, loss ratios for a particular customer group that were showing signs of being systematically higher than expected.  He gave some recent examples where Mr Godfrey had made such requests although mostly these occurred in 2014 outside the Period.  We accept, however, that the business operated essentially in the same way during the Period. 

Rate change proposals

234.       Mr Gumbrell said that the team made proposals for changes to the level of net premium for particular segments of the book, in response to risks or opportunities that had been identified as set out above, or as part of the implementation of pricing changes set in the Advantage budget, when that was introduced in 2014.  The pack of information provided to attendees of the RAC meetings also included any such proposals with supporting evidence and rationale for discussion at that meeting.

235.       Mr Gumbrell described the objective of the team’s work, as regards changes to the net premium, as being to enable Advantage to set them at a level with the right balance between net premium and loss claims so that risks were priced correctly and different groups of customers were not cross subsidising each other.  The team’s role involved balancing risk.  Net premium changes were not made by considering commercial factors, but rather by looking at risk pricing. 

236.       The process did not involve, therefore, considering the correct level at which to set the appellant’s commission, which was set by the appellant’s retail team.  For example, Mr Gumbrell’s team would not propose that Advantage should reduce net premiums in order to assist the appellant to hit its volume targets if reducing the net premium would increase Advantage’s loss ratio.

237.       There was, however, an exchange of information between the appellant and Advantage on their respective pricing plans and the effect of pricing changes as both Mr Gumbrell and Ms Johnson explained.

(1) Proposed net premium changes were provided to the appellant’s retail team as they could impact on that team’s commercial decision on the optimal level of commission. 

(2) In addition to providing information on the effect of gross premium on volumes and the mix of business (see [231]), the retail team also provided information on the relative competitiveness of Advantage’s net premium rates, suggestions for areas where Advantage may want to consider reducing its net premium rates and the impact of net premium changes on the appellant. 

(3) Mr Gumbrell noted that communication with the retail team also helped to check for unintended consequences of rate changes which could affect customer service, such as an increase in call centre demand, as the appellant had information on the impact of changes in the market. 

238.       If Advantage raised a query about the impact of the retail team’s activities on Advantage’s book, Mr Gumbrell discussed this with Ms Johnson on Advantage’s behalf.  Mr Gumbrell noted that neither his team nor the retail team had any sign off rights in relation to net premium changes.

239.       Mr Lee summarised the process leading to a rate change proposal as involving (a) the initial analysis of a particular area (such as the loss ratios of drivers of a particular age in a particular postcode) (b) the creation of a performance matrix that included all relevant loss ratios and inserted pricing proposals next to those ratios and (c) an analysis of the performance matrix in the light of how the team thought the loss ratio would develop in the short, medium and long term.  In the short term the team considered primarily the impact on sales balanced against the long term impact on loss ratios. 

240.       Mr Gumbrell described the process in more detail.  He said his team took the forecasts of claim frequency and claim severity for particular customer groups from the complex risk models prepared by the appellant’s actuarial team to forecast pricing, and reconciled the predictions against the data they gathered from their ongoing analysis of the performance of Advantage’s book.  A proposal for a rate change was, therefore, based on the appellant’s view as to whether the net premium level was sufficient according to (a) what type of claims had been made, (b) the frequency of claims and (c) the market data.  He noted that, however, when a policy was sold, it was not possible to know what the costs arising from the policy would be until, in some cases, as many as six years later.  Whilst a lot of complex analytical work was carried out by the actuarial team,  as his team had to work on a “best estimate” view of the costs to fit the actuarial predictions to the development of claims, there was necessarily a degree of subjectivity in their analysis of how Advantage’s loss ratios would develop.

241.       For example, in relation to theft or accidental damage perils, claims frequency could be relatively readily converted into a rate of loss because Advantage’s exposure was probably limited to the value of the vehicle.  However, for personal injury to third parties, claims frequency is a much less certain indicator of the likely overall exposure given the time these claims typically take to mature.  In those cases the team would look for other indicators of performance such as how Advantage’s premiums compared with others in the market.  If they were a lot less it was likely that Advantage was not collecting enough premium and may ultimately have a higher than expected loss ratio.  In forming a view about the performance of Advantage’s loss ratio through this complex process the appellant’s staff had to make a judgement, as did Advantage’s staff, when deciding whether or not to make the proposed change or whether further analysis was required.

Changes to risk acceptance criteria

242.       Proposals for changes to the risk acceptance criteria, with supporting evidence, were also put to Advantage at the RAC.  The setting of risk criteria was important in helping Advantage manage their reinsurance risk in terms of achieving the balance of risk which accorded with Advantage’s risk appetite. 

243.       Mr Gumbrell noted that, although Advantage’s reinsurance programme limited its exposure to large claims, those claims were still important; they could lead to volatility in the performance of Advantage’s book as a single claim or small number of claims may have a material impact on the overall loss ratio and could increase its reinsurance premiums.  He said that there was an element with large claims of “just bad luck”, which was unavoidable.  But there were also elements of judgment that Mr Gumbrell’s team tried to take into account in proposals to Advantage. For example, if they identified risk factors for a subset of drivers that had a particularly high propensity to give rise to large claims, Advantage could change the risk acceptance criteria so that this particular subset of driver was not allowed on cover.

Presentation of net premium/risk criteria changes at meetings

244.       A member of the appellant’s team presented the proposal for a rate or acceptance criteria change to Advantage at the meeting and it was discussed typically by Mr Charlton or Mr Godfrey questioning and challenging the analysis.  Mr Gumbrell said that quite often the challenges were made because the proposal contained a number of changes and, for example, Mr Godfrey wanted to be satisfied that the appellant had considered properly how the changes may interact and whether there were any alternative approaches which would give the same result.  He noted that Advantage may ask for further analysis and/or may want to see more evidence supporting a proposal. 

245.       Mr Godfrey said that, in advance of the meeting, he looked at how Advantage’s book was performing from the information provided by the appellant’s team, in particular, as regards the data in relation to the budgeted gross written premium and loss ratio.  He asked questions on the proposals presented at the meetings based on that knowledge.  For example, if the information showed Advantage’s performance was in line with its budget and the appellant proposed a reduction in net premium rates, he would query why the change was needed.  If the loss ratio of a particular segment of the book/category of customer was higher than expected, he would expect the appellant’s team to propose a rate increase, unless they explained why they believed the rate should be held or reduced.  On occasions he may query whether a proposal should be altered to take into account another “dimension” such as the driver’s age, postcode or the age of vehicle.

246.       Advantage’s staff had full access to the reports and the data analysis undertaken by the appellant, if they wished to verify a specific item.  Mr Godfrey noted that he and Advantage’s senior actuary had access to the appellant’s computer system and regularly (on at least a weekly basis) reviewed the underlying loss ratios and policy profiles.  They sometimes examined the excel spreadsheets on which the appellant’s proposals were based, when that was considered necessary to validate the work on the underlying analysis.  Mr Morales worked with the appellant’s team on the claims cost models and actuarial analysis.

247.       Mr Godfrey and Mr Charlton both said it was Advantage’s decision whether or not to adopt the proposals or to require the appellant to undertake further analysis which may require further discussion at a further meeting prior to a decision being made.  All presentations were reviewed in detail and they included impact projections so there was a clear understanding of the expected outcome from the proposed change.

248.       Mr Godfrey said that, following this process, there was not normally any disagreement on net premium pricing, although, on occasions, for strategic reasons (such as to meet profit and loss targets) Advantage did alter the appellant’s proposals.  In addition, Advantage sometimes proactively requested rate changes if they were concerned that the targets in the budget may not be attained.  He thought that the appellant and Advantage rarely, if at all, had major disagreements in relation to the rate changes because, although there were details of the changes to be determined, the broad parameters of net premium rate changes were set out in the Advantage annual budget.  The appellant’s underwriting support team were aware of these targets.  If the financial results were not developing in accordance with the budget, then Mr Godfrey would make a proposal for an adjustment to the broad net premium rates and request that the appellant provided specific proposals at a later meeting.

249.       This budget process commenced in January 2014 and prior to that Mr Godfrey made decisions regarding broad net premium changes at the meetings on a more ad hoc basis according to his view of what was required to comply with Advantage’s budget.  Mr Godfrey noted that the appellant could propose that Advantage did not go ahead with a budgeted rate change, for example, if the loss ratio was on track but gross written premiums were below budget but it was always Advantage’s decision whether to accept that or not.  The appellant had to implement whatever was decided. 

Challenges to proposals

250.       Mr Godfrey, Mr Lee and Mr Gumbrell all said that it was common for the appellant’s proposals to be challenged by Advantage and, as Mr Lee said, to a lesser extent, by the appellant’s senior executives (who generally had an opportunity to discuss any concerns prior to the meeting).  Mr Lee said that every net premium rate change was challenged by Advantage and around 50% of proposals were rejected or modified by Advantage. 

251.        Mr Lee said that a challenge may involve asking the appellant’s analyst to undertake further analysis, such as preparing a performance matrix that projected the development of the loss ratio over longer periods.  The challenge would typically apply experience to the mathematical data by, for example, requesting that a cell is split into another “dimension” (such as the postcode, the age of the car, the age of driver etc) or to postpone the charge whilst it was seen how the loss ratio developed. 

252.       Mr Gumbrell noted that substantial underwriting experience was key to the balancing approach needed to analyse the risks and the subjective decisions about likely future claims trends and development, or to assess the likely impact of proposed changes.  In his view, both Mr Charlton and Mr Godfrey were, given their extensive experience, “well-positioned to challenge our rating and acceptance criteria proposals and the basis on which we have formed them”. He noted that Mr Morales role was similar in that he made challenges on ways to improve proposals.  He was able to do this because part of the evidence for a proposal made by the appellant was the predicted loss ratio and this was produced by the actuarial function which Mr Morales oversaw.

253.       Mr Gumbrell gave the following examples of occasions on which Advantage challenged rate change proposals recommended by the appellant’s underwriting team:

(1)  In November 2014, the team proposed an increase in rates for high value vehicles because of high loss ratios.  The team wanted to increase net premiums pending further investigation of the matter.  Mr Godfrey disagreed as he thought customers with such vehicles were relatively low risk given they typically have a lower propensity for incurring personal injury claims.  Also, he was concerned that monthly sales to such customers was low which suggested Advantage was not under-charging those customers.  Advantage requested analysis as to why the claims costs were higher than anticipated when the net premiums were set, in order to consider whether the claims handling process with those customers could be improved. 

(2) In December 2014, the team proposed to decrease rates for the motorcycle account to take effect in January 2015 with a view to increasing the volume of policies sold to bring the account closer to its gross written premium target.  Mr Godfrey rejected the proposal on the basis that the change would be implemented out of season (the motorcycle season runs from around April to September each year), and therefore there was a risk of attracting a larger mix of commuter risks that use motorcycles all year round in all weather conditions, resulting in higher exposure and frequency of claims. Advantage also raised as a concern the risk of the reaction of competitors in the market.

(3) In October 2014, the team proposed to hold off changes in renewal rates on a particular account in order to support retention rates that were below target.  Mr Godfrey disagreed with this recommendation and budgeted rate changes were applied across both new business and renewals.

(4) In June 2014, the team made a proposal in relation to commercial vehicle rates that was challenged by Advantage. Advantage considered that further analysis was required in order to establish that this was a genuine trend.

254.       Mr Godfrey said that he rejected a proposal in October 2014 to hold back on a budgeted rate increase because a new rating system was being brought online for renewal business.  The appellant wanted to wait and assess the impact of the system change before implementing the rate change.  However, Mr Godfrey said that the budgeted rate change should be introduced as originally planned because claims ratios were still uncertain.

255.       We note that the examples did not relate to the Period.  However, we accept that they indicate how the procedure operated during the Period given the evidence is that such challenges were also made in the Period and these are illustrative only.

Importance of risk selection

256.       Mr Lee agreed that the analysis done by the appellant as described above was very important to Advantage and to the group.  The analysis directly concerned the profitability of Advantage’s business, and also, more generally, the group; if the appellant got the risk analysis and rate change proposals wrong, the risk was mis-priced and Advantage stood to make less profit or even a loss.

257.       It was put to him that the appellant’s team was doing sophisticated risk selection which the RAC could then assess.  He said the data analytics were extremely important; the customers filled in 30 or 40 questions about themselves and the appellant enriched that with other data.  Pricing proposals that used that level of sophistication and data enrichment were presented to the RAC for them to look at it from “their different lens, as to the selection of the risks that they wanted to write to maximise the value that they got out of the capital that they were managing”.  The risk selection was made once Advantage had agreed and set its structure. 

258.       He continued to explain that there were multiple tables that were put into a rating engine, and within “a millisecond”, that computed the premium taking into account all the factors.  So risk selection was not done on an individual basis as such.  It was the selection of those tables and the multipliers within those tables for the different risk factors that Advantage decided upon, based on data that was presented to them from the appellant who had the expertise in that type of data analysis.  But it was the experienced insurers at Advantage with a “different lens on”, who looked at how changes that could be made to the portfolio, to the average premium, could affect Advantage’s capital position, its reinsurance structures and compliance with the regulatory insolvency requirements.  So what may be termed the risk selection, in his view, was done by Advantage setting those tables and pricing structures and rules. 

259.       Mr Lee noted that whilst Advantage had chosen to employ the appellant to do the analytics for them they could equally employ Ernst & Young or Deloittes or many other specialists in that particular field.  He agreed that the appellant’s role was very important because, without the sophisticated analysis, the directors of Advantage could not exercise their judgment and do their job.  He agreed Advantage certainly would have to employ somebody to do that, as they were making their decisions based on the quality of that analysis but, as he said, there were many other companies that would do this for Advantage if it chose to use them.  He thought that Advantage used the appellant because they considered the appellant to be the “best”. 

Other changes

260.       A similar process was followed in relation to work on specific initiatives, such as the project which assessed the rate changes required to comply with the Gender Directive 2004/113/EC which required insurers not to discriminate between customers on the basis of their gender. Mr Morales managed the project through weekly meetings with members of the appellant’s underwriting team.  The majority of analysis was carried out in Bexhill but Mr Morales instructed the Bexhill team as to the actions required and the methodology to be used and provided insight and knowledge of the market.  Had Mr Morales not been involved in running the project, different software and techniques would have been used to carry out the analysis, as the Bexhill team were not familiar with the particular software he directed it to use at the time.  One part of the project relating to changing the acceptance rules to become gender compliant was delivered solely by the Bexhill team with little input from Mr Morales but this was a small component in the context of the entire project.  Mr Charlton managed a telematics project in a similar manner.

Pricing – validation of changes

261.       Once a rating decision was approved by Advantage, the appellant’s underwriting support department implemented it by submitting updates to the tables of pricing rates and acceptance criteria that were held on CDL and conducted tests to ensure the change was working in accordance with the specification provided.  They could only make such changes on receipt of such an instruction.  As noted, Advantage’s underwriting team also monitored the outputs of the pricing system.

262.       Mr Gumbrell explained that the appellant ran various systems checks to ensure that a rate or risk acceptance criteria change had been implemented correctly and that it did not cause unintended changes elsewhere in the systems.  The appellant “batch validated” 50% of net premiums for private car policies. The validations were weighted around situations where there was greater risk.  For example, on days where there was a change, the appellant may validate 100% of new business sold.  The appellant’s testing team created a test plan based around the type of change to ensure it would be delivered on CDL in accordance with what was intended.  The test plans had to be approved by Advantage to ensure that they were adequate and were focussed on the areas that had the highest risk of error.

263.       Mr Gumbrell said that Advantage reviewed the validation reports compiled by the appellant’s testing team.  If the reports showed that there were cases that did not reconcile, Advantage asked for more information or undertook work to quantify the size of the issue.  Advantage itself performed all the validations on motorbike and commercial vehicle policies against a spreadsheet maintained by Mr Powell.  Although motorcycle policies were a small part of Advantage’s business, they resulted in a disproportionate number of large loss claims and were therefore important in terms of risk.

264.       Mr Gumbrell said that he would expect the audit and validation work that Advantage carried out to be carried on by an insurer as an insurer would want to ensure the premiums were being calculated correctly on the brokers’ trading platform.  If Advantage did not have the resources to do this in Gibraltar, he would expect the validation to be carried out in the UK.

Pricing – gross premiums

Overall approach to setting the gross premium

265.       Ms Johnson expanded on how the appellant’s retail team set the gross premium.  She noted that the total commission earned depended on the number of policies sold and the commission per policy.  Setting the gross premium was a question of achieving a balance to find the right point between charging more commission and potentially getting less customers, and charging less but getting more customers. 

266.       Essentially the team faced outwards towards the market and assessed how much broking value could be obtained from a given customer and how “price-elastic” that customer was.  So, for example, it could be that an insurer had some specific information about the particular customer that enabled it to provide a cheaper net premium quote for that customer than that provided by other insurers.  In that case, the appellant would know that the customer’s expectation of price was higher and could use commission to fill that gap and generate value for itself.

267.       In the past some of the appellant’s insurers paid the appellant only a fixed percentage commission.  That meant it could flex the final price to the customer only through giving a discount in its commission.  The appellant has moved away from this to setting its own commission so that it has more flexibility over pricing.  In all new requests for proposals received from insurers since 2013 the appellant has had that ability. 

268.       Ms Johnson said that the market also had to be factored in as a very important element.  The appellant did not operate in isolation.  If it wanted to drive a great deal of volume in terms of policies sold, it could reduce prices, but that could cause the market to react and actually destroy value.  Prices would be driven down and volume in the end would return back to the previous level but everybody would be making less money.  She also noted that the appellant considered a variety of customer values, of which commission was only one (as set out in further detail below). 

269.       Although the team’s aim was to maximise the appellant’s profits, the team was also concerned with ensuring that the correct business mix of risk was underwritten.  The underwriters may request business mix constraints (such as restrictions on the proportion of young drivers) due to constraints under reinsurance contracts (see below).  She explained that: 

“our systems will sell to the customer for the optimal price that retail set.  We will sometimes receive feedback post-sale to say that business mix is less favourable to a particular insurer…… And in which case we will be manually keeping an eye on that mix of business, and if needs be we can make commission adjustments to ensure we do not exceed particular thresholds that underwriters might have”.

270.       Mr Pavey similarly noted that the potential effect on the insurer influenced it in how much commission to charge.  He said that on occasions the appellant may provide a discount for the customer, by charging a gross premium for the policy which was lower than the net premium.  However, that might result in an insurer writing more business into a particular market segment than the insurer would expect or be comfortable with.  The insurers could analyse the segments of the market written by the appellant and its commission rates for their policies and could monitor the loss ratio performance of their book of business.  The appellant’s business as broker would be adversely affected if insurers were not satisfied with the business mix. 

271.       Ms Johnson said that, however, the ability to set its own commission means that the appellant can use the level of gross premium to affect the volume of policies sold, when it has a business need to do so.  She gave an example of when the appellant had done this.  In the summer in 2014, there was a staffing issue in the appellant’s call centre, such that it was falling behind in answering calls.  A decision was taken by the appellant to remove its telephone numbers from aggregator websites and to increase the gross premiums to reduce call volumes so that the call centre staff could work through the backlog of calls and maintain a good standard of customer service.  The appellant made this change across all of its product lines and in relation to all panel members.  Prior to this change, the appellant notified all the panel insurers that this would be done, but approval was not sought. The notification was mainly given to them to keep them informed about the impact on the business that they would be underwriting during that period.  They were provided with information on the difficulty, how long it might take to resolve the difficulty and the proposed actions.

No input on net premium

272.       Ms Johnson noted that the retail team did not sit on the RAC and had no approval rights as regards Advantage’s net premium rate changes although some of the team had discussions with Mr Gumbrell’s team, at least on a monthly basis, regarding the matters set out below.  The team did not have any input into the net premium insurers inputted into the CDL system and any rating changes made in that system.  As for all its insurers, however, the team managed the process of making changes through RTP on receipt of a formal request by the insurer.  As noted, the rules determining the gross premium were also held within RTP and were applied after the insurer rules.

Determining the gross premium

273.       The team set its own “hard” and “soft” limits for possible changes to gross premiums.  The team controlled these limits, which were periodically reviewed and refreshed, according to commercial appetite and good customer conduct.  When changing the gross premiums through RTP every two to three weeks, the team did not allow the commission to be adjusted by more than a certain percentage of the gross premium in any one change.

274.       Gross premium price changes above a certain threshold had to be referred to the executive committee, although Ms Johnson was not aware of such a referral having made.  At the hearing she said that she thought the level where executive approval was required was around a 5% increase.  An increase of that size could cause a market slide such that it was prudent for the executive committees to have sign-off.  That was the only point at which Advantage would have any input on price changes.

275.       In determining the gross premium, the retail team worked out a “customer value” by reference to the income streams associated with the policy multiplied by the length of time that the customer retained the policy. The relevant income streams were (a) the appellant’s commission (b) ancillary income from any associated products such as breakdown or legal cover products (c) premium finance income, being the interest paid by customers who paid by monthly direct debit, and (d) fee income (such as set-up fees or mid-term adjustment fees).

276.       Working out the optimal gross premium given a fixed net premium involved assessing a number of factors including the customer’s price sensitivity, whether the customer would pay by direct debit, the customer’s propensity to buy ancillary products, to cancel the policy and to renew the insurance policy and the underwriting value of a policy (if underwritten by Advantage). 

277.       In undertaking this exercise, the team looked at: 

(1)  The “broker lifetime value” - the expected value of the policy holder to the appellant, as broker/intermediary, taking into account the customer’s expected propensity to retain and renew their policy.

(2)  The “group lifetime value” - the expected value of the policyholder to both the appellant, as broker/intermediary (on the same basis as in (1)), and Advantage as insurer, taking into account the expected loss ratio for Advantage.  For car insurance underwritten by Advantage, the team used systematic price optimisation processes, which predicted the loss ratio for the policy, and used that, together with predictions regarding the appellant’s retail income streams.  They used a less sophisticated process for determining the customer value for all other policies.  The team received occasional information from insurers on how their loss ratio was developing and discussed with them how this was likely to impact their net premium rates.  We note that Mr Pavey confirmed that, as regards Advantage, only the broker lifetime value was taken into account until around July 2011.

(3)  The “retail income per policy” - the single year view of income for any given transaction assuming no cancellation of a customer, which would take into account some or all of the factors described above.

278.       Commission price optimisation for all other areas involved the pricing team proposing commission changes, assessing how these changes would impact volumes (using elasticity calculations) and completing a pricing proposal.  This would usually look at one to three year group and broker lifetime value. 

279.       For commission changes for insurers other than Advantage there was no systematic consideration of the insurer’s underwriting profit although the appellant had conversations with them about their loss ratios and did care whether or not the business underwritten through the appellant was profitable.  The team worked with other insurers to understand their requirements for profitable business.  If they identified that the insurer’s book had particular areas of strength in terms of loss ratio (for example, if the loss ratio was low for a particular segment of customers), the appellant may, if it thought focussing on that area would complement the appellant’s business aims, introduce discounts to its commission specifically for that area.  In effect this may support the appellant financially, through increased retail income streams, via a form of profit share based on loss ratio, or just be worthwhile to ensure that the business written by the appellant for that insurer remained profitable so the insurer would continue to use it.  The appellant has begun to introduce limited profit share arrangements (some of which are already in place) with other panel members and so this consideration is growing in importance.

280.       The team took the decision on what price to sell the product for in order to maximise customer value whilst also considering volume targets.  It was for the retail team to decide at what price to sell an insurance policy.  It had independent targets for customer value and volume, although there was some interaction with Advantage’s targets for the business that it wanted to underwrite for the year.  The team considered the budget and analysed how much the market was likely to grow and develop in terms of pricing.  This was significant because if, for example, the market grew by 4%, all things being equal, the appellant’s volume should grow by 4%.  These types of assumptions were very important to the growth projections for the account volumes.

Exchange of information with insurers

281.       Ms Johnson confirmed that, as Mr Gumbrell had said, developments and changes to gross premiums were regularly referred to the appellant’s underwriting support team, so they could report changes to Advantage.  That information could help an underwriter understand areas of competitiveness in comparison with the market as a whole.  Similar information was provided to other panel insurers, but on a less frequent basis.  The difference reflected that more business was underwritten by Advantage and the appellant, therefore, wanted to maintain a strong relationship with Advantage. 

282.       Ms Johnson confirmed that Mr Gumbrell’s team were also notified of the account volumes’ analysis for provision to Advantage.  High level volume analysis was shared with the other insurers but with less detail on the assumptions behind the analysis.  As those insurers generally used a number of intermediaries to sell their policies, they placed less importance on the volume analysis than Advantage. 

283.       Mr Gumbrell’s team gave the retail team the projections for claims inflation which meant they knew whether they could expect Advantage’s net premiums to become more or less competitive in the market.  Ms Johnson said that this process should happen with other insurers but the level of sophistication of approach by other insurers is often lower.  On occasions the appellant found that an insurer may not have made appropriate assumptions as to the levels of claims inflation or price inflation.  Ms Johnson tried to challenge these assumptions (by reference to the appellant’s data and projections) to encourage the insurer to be more realistic about future volumes of business and/or pricing of risk. 

Sales of policies and managing the business mix

284.       Ms Johnson noted that, in effect, under the automated system the team ask the insurers to provide a net premium for a customer.  If a price is provided that means the insurer is willing to do business.  The team then has quite a lot of flexibility around which customers it chooses to do business with on the basis of quotes provided.  At the point of sale of the policy there is no intervention by the insurer.  As it is for the insurer to price the customer risk, the insurer needs to find the most effective way of judging that risk.  She said she would expect any of the panel of insurers “to provide us with a price that they are comfortable with for the risk for that customer and I would expect that price to also include a margin”.  She noted that the setting of the acceptance criteria is an important way in which an insurer manages its risk.  For example, an insurer may have policy rules such as not insuring people outside certain age ranges. The retail team “enrich that judgement by feeding back with external commercial information about the market”. 

285.       As noted, in fixing the gross premium the team had regard to insurers’ requirements in relation to the mix of business that they underwrote.  Advantage was more sensitive than other panel insurers to changes in customer mix because it only used the appellant as its intermediary.  Ms Johnson said that it was important for the team to check and correct the balance of business written.  The appellant did not want the insurers to manage business mix and other issues when they were setting the net premium.  The retail team had a better view of the market and, therefore, wanted to help insurers achieve the correct balance by taking steps to sell policies to the right mix of business, rather than the insurer manipulating the net premiums to try to adjust the mix. 

286.       The appellant provided the insurers with feedback, for example in relation to areas of the market where they had low “conversion” of customer quotes into sales and of areas where they were and were not competitive or if there was a big area of the market where they were doing no business. Insurers could then consider whether they were able to provide lower net premiums.  As the team’s focus was to drive sales, they were constantly seeking more competitive net premiums from insurers and, as noted, it was for the insurer to determine whether from its perspective the risk was properly priced. 

287.       The relationship with Advantage was the most important insurer relationship because Advantage underwrote around 90% of the appellant’s business.  Ms Johnson thought that this was because it tended to provide the best net premium price for customers, all of which business was then sold through the appellant as its sole intermediary. 

Exchange of other information

288.       Analysis that was sent to the underwriting support team included a weekly scorecard generally showing the trend over time which enabled business performance to be monitored.  The other panel members received similar data, such as sales volume and average premiums, but with less detail and on a monthly basis.  The appellant has invested less in data development for smaller panel members. The scorecard was a business wide analysis used to manage the weekly performance across all of the appellant’s areas: commercial, marketing and operations much of which was not relevant to Advantage.  Advantage used the scorecards to monitor the mix of business, the sales volumes and the loss ratio.

289.       The flows of information to other panel members were less sophisticated than the flows of information to Advantage, partly due to a difference in the systems that were available.  Other panel business was carried out in Newmarket, where the appellant only had sales data, whereas Advantage business was carried out in Bexhill, where there was the ability to compare sales and market data (which needed to be compared to provide information on how competitive an insurer is). (This is changing so that all new car business will be carried out in Bexhill).

290.       It was the agreed approach for both Advantage and other panel members to engage with the team early about underwriting changes, but sometimes sight of these changes was provided later than the team would have liked.  When Mr Gumbrell’s team had almost finished developing a net premium rating change proposal for Advantage, they shared the analysis in relation to (a) levels of net premiums, (b) customer retention rates (c) budgeted rate increases and (d) information relating to the market (for example, information projecting what the effect of an increase in the net premium would be).  The team provided feedback to Mr Gumbrell’s team in relation to how the changes to the net rates were likely to affect the appellant’s trading performance (such as volume of sales, income per policy, gross written premiums etc.).

291.       Ms Johnson gave an example of interaction with Mr Gumbrell’s team where that team undertook some work on the reclassification of vehicle groupings for its private car customers, which lead to Advantage making a net premium change.  The retail team were informed of this change and were given an indication of its likely impact.  The changes made impacted both on the appellant’s ability to provide quotes for Advantage and the competitiveness of quotes.  As soon as this became clear, the retail team reported the impact to Mr Gumbrell’s team to enable them to review the impact of the adjustments and to discuss with Advantage whether the desired result was being achieved.

292.       Ms Johnson noted that sometimes, changes made by Advantage to renewal net premium rates caused the appellant difficulties in that they considered that, commercially, the increases were excessive.  The team had no decision-making rights but, having looked at Advantage’s net premium and the loss ratio compared to the budget, they would debate the rate changes with Advantage through Mr Gumbrell.  Disputes could be escalated to the group’s executive committee but disagreements had, in Ms Johnson’s experience, always been resolved without the need for formal resolution by the executive committee.  The result of these debates could be that the net premium change went ahead anyway (if Advantage needed to protect its loss ratio), but with more clarity as to the consequences of the change, or it was modified or the implementation delayed.

293.       This analysis of the net premium rate from a commercial point of view was also carried out in relation to other panel insurers.  For example, as regards home insurance, the team noticed that one of the insurer’s net premium rates had gone up about 12% year on year rather than 2 to 4% as had been predicted. This was reducing the retention rates, so the team discussed this with the insurer.

Pricing – further evidence on interaction between net and gross premium pricing 

Appellant’s evidence – Ms Johnson, Mr Lee and Mr Gumbrell

294.       It was put to Ms Johnson that there were tensions between the appellant’s desire to maximise its commission and Advantage’s desire to increase its volume of sales.  She said that setting the commission was based around making sure that the price was appropriate for the particular customer given the prices in the market. The appellant tried to be very competitive but not first.  The team did not use commission “as a way to steal a bunch of type of customers”; they used it as a mechanism based on how likely customers were to buy.  Some of the customers with the highest amounts of commission, such as those aged over 50, may have high commission because the appellant needed it to make that category of customers profitable for it.  However, it could well be the case that the appellant was nevertheless successful in attracting those customers because it could still provide a good price versus the rest of the market. 

295.       She said that the insurer needed to have a profitable broker and part of the key profit stream for a broker was commission.  An insurer would not like a broker to make excessive commissions.  If volume was particularly low the insurer may query whether the appellant was doing anything to constrain it.  She thought that in general, however, there was a good understanding that commission was an accepted and important income stream for a broker.

296.       It was put to her that Advantage was vulnerable, as far as the gross premium is concerned, because it only had one intermediary.  She said that on the other hand you could say that Advantage was very focused on writing really competitive prices and had a broker which had some competitive advantage in the marketplace, which combination was proving to be very successful.  If it was writing business with many intermediaries/brokers, they would be competing against each other.  That may work for some as a business model, but the Hastings/Advantage model was proven as being a very strong business model too.

297.       She noted that it was not in the appellant’s interest to simply charge a very large premium.  If it did so it may earn lots of commission in year one but, due to the high premium, the likelihood of those customers cancelling or not staying with the appellant in two and three years’ time would be extremely high.  So that approach would be a very short-term one, even just from the broker’s perspective.  Ms Johnson would not recommend, unless there were some extenuating circumstances, that prices were “hiked” and volumes decreased in that way; it would not make long-term value sense.

298.       She agreed that there could be circumstances where there was some conflict of interest between the appellant and Advantage, in terms of their respective precise strategies to earn profits.  In those cases of what she described as “healthy tensions”, the appellant would lean towards supporting Advantage as regards its loss ratio because that was important from a business sustainability perspective.  The appellant did not want to write business where insurers risked their loss ratio, as that would be short-termist.

299.       Mr Lee said Advantage was interested in the gross premium essentially with a view to managing the capital requirements and the reinsurance provisions.  At a “micro level” Advantage wanted to know how profitable the broker was.  He noted that primarily the underwriter created a net rate based on its expected claims losses, plus a margin.  But underwriting motor insurance is a cyclical business and there are times where it is more or less profitable, and if Advantage saw that in certain segments the appellant was taking a lot of commission:

 “they might just put up their rates just to take some of that profit, and say, “I do not want the retailer taking all that additional profit”…..And they therefore just put a net rate increase in, not because they need it, not because the loss ratio is deteriorating, but because it is available and they do not want the retailer to take it all.  And that occurs all the time, in good conditions”.

300.       In re-examination he confirmed that would be the same case as regards other insurers if they operate on a net premium basis, as most of them do.  If they saw the broker was taking excessive commissions, they would want a share of that.

301.       It was put to him that it was because the gross premium affected Advantage as well as the appellant that major changes to commission were subject to approval at the executive committee.  He said that the net rate was “hugely important” to the appellant, but the parties did not make each other’s decisions.  He said that “ultimately if Advantage was not comfortable with the commissions that Hastings were putting on the policies, it could choose to move distribution elsewhere…very easily, as it had in the past”.

302.       It was put to Mr Gumbrell that Advantage was interested in the gross premium because, given that it only had one distribution channel, an increase in commission would decrease its sales of insurance even where Advantage had the capacity to take those risks on.  He said:

“It could do.  I mean, it depends on the environment relative to the rest of the market.  If Hastings chose to increase commissions above and beyond competitors it could have that effect.  I guess, if it got to the point there is…..nothing to stop Advantage approaching, going back onto other panels, like they used to with Kwik-Fit or at the AA or someone else who would be prepared to accept or write business at a much lower commission, which is what used to happen”.

303.       It was put to him that there was collaboration between Advantage and the appellant as regards the gross premium targets.  He said he thought the appellant had to discuss the gross premium with Advantage “because the gross premium has two components, the net rates and the commission.  So naturally both have a shared interest in that.  That is the same with any insurer/broker relationship”.

304.       He continued that the appellant producing the management information for Advantage was not really about how many policies were sold and how much gross premium charged as:

“We are not looking at commission - that is a Hastings function.  It is their job to look at commissions and decide what the right level of commission is to apply, given all the things they need to do to deliver income for Hastings.  Certainly when we are making recommendations to Advantage we are not considering commission.  We are considering primarily loss ratio and secondarily volume or total premium”.

305.       It was put to him that both parties had an interest in each element of the total price.  He said that:

“it would, in the same way it would with any other broker/insurer relationship, yes.  As I say, the gross price is made up of the commission and the net rate.  When we are going through the budgets process, the real concern is that…we are predicting the claims costs…. the net premium is enough to cover the claims cost and deliver whatever profit target is for Advantage.  What Hastings then does with the commission levels on top of that are down to [the retail team].  It will have an impact because, you are right, the price the customer sees is a combination of both things……But first and foremost Advantage’s main concern is to achieve the loss ratio target, and it would not do that at the expense of writing more policies.”

306.        He concluded again that if the appellant raised the gross premium in a way that thereby depressed customer demand overall for Advantage’s policies, such that Advantage was not using its full capacity to take on risk, then “at that point….. Advantage would be looking to potentially sell business through other broking channels where they could do so and with the lower commission.  But that is a decision Advantage would have to make”.

307.       It was put to him that the gross written premium directly affected the budget because it would determine how many policies overall were sold.  He said:

“The elements that we were effecting of that gross written premium target were the average net price the customer pays and the number of customers.  There were things that Hastings, as the broker, did that would also influence the number of customers that we would sell - for example, if they changed something on their website or ran a marketing campaign or whatever else, that would also influence it.  All of those things would have to be taken into account in the plan.  It would be wrong not to consider all of those things.  And they all go into the gross written premium target”.

Advantage’s evidence – Mr Godfrey

308.       Mr Godfrey acknowledged that “if Advantage puts its net rate up….you would think it would be less competitive in the market, unless the market has put its rates up by the same amount”.  It was put to him that meant there were diverging interests between Advantage and the appellant.  If Advantage increased the net premium, it would be more difficult for the appellant to sell Advantage’s policies and it would earn less commission, whereas if Advantage decreased the net premium that would potentially benefit the appellant.  He agreed but said:

“if we needed to put net rates up because of claims performance, then that is what Advantage would do. And then, for that particular segment…..the appellant would be less competitive, unless, as I say, the market….. put the rates up at the same level”.

309.        It was put to him that this meant there was scope for disagreement on net premium rate changes; whilst Advantage might want to increase rates because of concerns about the risk that they were undertaking, the appellant was focused on earning commission.  Mr Godfrey said that was why the business was structured as it was, “so that Advantage had control of net rates”. 

310.       It was put to him that the reason why, as he said, there were no major disagreements on the net premium changes must be because the interests of the two companies were aligned.  He said that actually it was because Advantage generally had a rating plan for the year which it followed.  So the decisions generally were around how to implement those rate changes (as signed off by the Advantage board) rather than the direction of the changes.  But in any event he did not recall any rating discussions or decisions going on at the executive committee. 

311.       He said that it was quite clear what the structure was and the net rates were controlled by Advantage.  The underwriting support team at the appellant made their recommendations based on analysis and any trends they were seeing, and Advantage approved it or not.  It was a separate process in relation to setting the net premium and Advantage left the appellant to apply their commission:

“pretty much whatever they want to do with that [the appellant’s commission].  We are not bothered about that.  And if it slows down volume it slows down volume, if the rate has to go up.  So there is a clear separation on that”.   

312.       It was put to him again that the reason there was rarely, if ever, disagreement was because Mr Hoffman was only interested in the group performance.  If, overall, Advantage did not want to underwrite too much risk, and the overall performance of the group was thereby better it was not a concern that the appellant had less business.  He replied that would mean the appellant would not grow in the same way but Mr Hoffman might go to Advantage and ask for information regarding what was going on and the reasoning behind decisions:  

“But as I say, we have never discussed rates at the executive committee level.  That is done outside of that, and Advantage, as I say, have the say in what happens with the net rates.”

313.       Mr Godfrey agreed that Mr Hoffman took an overall view but “he would also be interested if Advantage was not competitive and…….was not meeting budgeted targets in terms of the premium volumes and policy volumes”. 

314.       He agreed that Advantage looked at the gross premium in terms of the profit and loss account and its financials.  It was put to him that Advantage must be concerned with changes in gross premium as its only distribution was through the appellant.  Mr Godfrey agreed that was why the projected gross premium was inputted into the Advantage budget which was then approved by the Advantage board of directors.  He thought that process would be the same whether Advantage were writing business through the appellant or, for example, the AA:

“We would want to know what their plans are in terms of what they are going to do with their commissions or whether they have any other initiatives that are going to generate more or less volume, so we can build that into our plan……So we look at what we think is going to happen in terms of the risk price, and model what we think is happening to claims inflation for claims severities and for claims frequencies.  But we also need to know what is likely to happen with volume and what the intermediary is going to do and what their plans are….they have information about the aggregator price comparison website.  They might have a view on what they are going to do, whether they will grow or shrink or whatever.  And that would be built into their plans as well.

So in terms of trying to project the level of business that Advantage writes, you need that element to it as well, because we need to know - that features in the planning.  And we need to know what capital we are going to need to support the business going forward. So yes, that is part of the process.”

315.       Mr Godfrey did not agree that the gross premium was set, at least, by reference to Advantage. He said:

“the gross premium is completely at the will of Hastings.  We do not input into what commission they are going to apply.  We want to know what they are going to do or what they are planning to do so we can build it into our plans, but we have no input into what level of commissions they charge.  And likewise, they do not have any input into what level of risk premiums we are going to charge.  But each other needs to know ….what the other is likely to do so they can plan accordingly…..we need to know what their plans are in terms of their gross premium so we can incorporate it into the Advantage budget that is presented to the Advantage board.”

316.       It was put to him that the Advantage finance team would not simply accept a gross premium if it would cause them a problem.  He said again that there would be discussions to understand what each other’s plans were.  He agreed that Advantage could go to Mr Hoffman, for example, if the appellant was rebuilding the call centre system and had upped the gross premium to reduce the volume but Advantage had capacity to undertake risk.  Mr Godfrey commented that Mr Hoffman would probably say in such circumstances “well, we cannot service that business because we are putting in the new system and therefore we need to slow volumes down.”  Advantage would then have to adjust its budget accordingly.  Mr Godfrey emphasised that, whilst he could look at the volume of business that Advantage was underwriting, which may tell him indirectly what the appellant had done in terms of the overall level of business, he did not know what they were doing on the commissions.  “As I say, it is completely separate, and not something that I am interested in.  I am interested in the net price and the risk price.”

317.       He agreed that, as with the net premium, the process of determining the most appropriate gross premium involves effectively a collaboration between the appellant and Advantage with the common aim to generate profitable business.  It was put to him that Advantage may have to accept gross premium increases because of particular issues that the appellant has, such as, its capacity to deal with the number of callers at the call centre.  He said:  

“yes, they are two separate businesses, and Hastings has their profit targets and Advantage has their profit targets.”

318.       In re-examination he clarified that this sort of collaboration was no different to the collaboration that might take place between Advantage and a third party broker or the appellant and a third party insurer.  It would be “the same, same considerations and, I guess, the same process”.

319.       It was put to him that just as with the net rate increases, Mr Hoffman must have been interested in the group profit figure.  So if Advantage went to complain to him about a large rise in gross premium he would not be worried by that loss to Advantage and would regard it as something Advantage simply had to accept.  He said:

“yes, from that perspective, if that affected the volume then we could….control our net rates and that will be the same.  As I say, if we were writing business with the AA or Budget or anyone else, if something is happening to them that is affecting their capabilities and volumes, then we would be affected by that”.

320.        It was put to him that in fact there were no major disagreements because the businesses were operating as an integrated model.  He said he never saw what the appellant was doing on its commissions.  It had:

“a team of guys, decision scientists, I think, that do a lot of their work on that, and it takes into account a number of factors, but I am not particularly interested in them……I never see a report of  when they change it.  I do not know what they are doing with their commissions. As I say, I am not really interested.  I can look at the volume of the business that Advantage are underwriting, which may tell me indirectly what they have done in terms of the overall level of business, but …..I do not know what they are doing on the commissions.  As I say, it is completely separate, and not something that I am interested in.  I am interested in the net price and the risk price.

321.       In re-examination he agreed that the motor insurance market is a very competitive market.  He said there are “cycles of rate increases and rate reductions, depending on how the market is performing.  And it is very competitive.  And the price comparison websites have probably accentuated that as well”.  He said that market forces or market constrains were a key factor in influencing the level of net premium that Advantage is able to charge, or indeed gross premium that the appellant is able to charge.

Claims handling

Claims handling guidelines – scope of the appellant’s delegated authority

322.       There was an essential distinction between the handling of large loss claims which had to be reported to Advantage for it to take the relevant decisions and all small claims, which were dealt with by the appellant under its delegated authority within the framework of the claims handling guidelines.

323.       Mr Eagar noted that large loss claims constituted approximately £30 million (gross) out of a total of £350 million of claims (gross) reserved on Advantage’s book and represented approximately 250 cases out of a total of 72,000 open cases. These were “long tail” claims which could take quite a few years to settle, therefore, underwriters and actuaries took a keen interest in them.

324.       Mr Lee noted that large loss claims were also examined individually as part of the quarterly reserve reviews which Advantage undertook for its external actuarial reports, and the actuarial reviews prepared by the appellant’s actuarial team for Advantage’s actuary.  The reserves also factored in the bulk data on the small claims relating to accidental damage, third party property damage and bodily injury.

325.       Mr Lee said that the guidelines were arrived at by a joint process between the appellant and Advantage.  They were kept under review and were updated from time to time (as in the increase of the large loss threshold in August 2010 and 2014 in order to keep up with claims inflation).

326.        He noted that, although the appellant was closer to the day to day handling of claims, Advantage was involved because claims handling impacted on its profitability and affected compliance with regulatory requirements and reinsurance contracts.  Advantage remained involved to some extent even in relation to small claims and, in particular, scrutinised the average monthly severity by peril and the appellant’s claims handling strategy.  These were discussed at the monthly ORM.

Dealing with reserving

327.       The claims handling guidelines set out the “reserving philosophy” for claims, which, as Mr Charlton confirmed, was set by Advantage, and was to provide for “any liabilities that have been incurred on insurance contracts as far as they can be reasonably foreseen”. 

328.        For large loss claims, a proposal for a reserve was made by the appellant in accordance with the philosophy.  However, the decision on the appropriate level of the reserve and on releasing the reserve was taken by Advantage following discussion at the LLC.  Mr Doidge noted that once Advantage had made a decision on a reserve for such a claim the appellant had discretion to amend reserves for the claims by a specified limit which, prior to 12 August 2014, was £10,000, where the total amount reserved exceeded £100,000.  Movements in reserves by more than that amount had to be notified to Advantage and proposals to move reserves by more than £100,000 had to be considered at the LLC.  Movements between the lower limit and £100,000 were agreed with Advantage as set out below.

329.       As both Mr Lee and Mr Eagar confirmed, reserving decisions on non-LLC claims were taken by the appellant under its delegated authority in accordance with Advantage’s reserving philosophy.  The appellant also took action to settle such claims under that authority.  For these claims the aggregate file reserves were reported to Advantage as part of the underwriting management information reports.  Mr Eagar noted that Advantage reviewed some of these claims as part of its on-going audit programme and could challenge the reserve as part of that process.

330.       In relation to any claim which was to be repudiated, the timing at which the reserve was to be released was decided by the appellant, if it was a non-LLC claim, and by Advantage (by the LLC), if it was a large loss claim.

331.       Mr Lee noted that the appellant’s anti-fraud work has enabled reserves to be released by Advantage more quickly than would otherwise have been the case.  This revision to the timing of the release of reserves was agreed with Advantage through discussion between the appellant’s counter fraud director and Mr Eagar.

332.       Mr Doidge said that, on the basis of this reserving philosophy set by Advantage, he had prepared detailed reserving guidelines for the appellant’s claims handlers to enable them to follow the reserving philosophy.  The reserving guidelines were circulated widely throughout the appellant and Advantage (and in particular to Mr Eagar, Ms McKeever and Mr Godfrey) and he consulted with them to ensure that they were all satisfied with the reserving guidelines and that it was an agreed document.

333.       Mr Eagar noted that the guidelines and reserve philosophy were important for large loss claims. The reserve philosophy formed part of the way the business calculated the amount of capital it needed to hold, how the actuaries reviewed that data and how that was passed on to the reinsurers.  The reinsurers in turn wanted to understand how claims were reserved so they could make sure that they had the finances to meet their eventual liabilities as set out in further detail below. 

Claims handling manuals

334.       As Mr Eagar explained, the appellant also operated in accordance with claims handling process manuals. These were compiled by the appellant based on the principles set by Advantage in the claims handling guidelines but they provided more detailed and easy to follow practical implementation guidance.  They were constantly being updated and evolved.  He noted that the appellant was required to have such fully documented claims processes because of its FCA obligation in the UK.  Advantage was likewise required to document and keep its internal procedures up to date for the FSC. 

335.       The manuals formed part of the means by which Advantage decided whether the appellant was handling claims in an acceptable manner.  Advantage reviewed them and any changes that were made as to whether they represented best practice.  Mr Eagar said that it was important for Advantage that, when auditing the appellant, it was doing so against the appellant’s own defined process manuals which Advantage had seen and was happy with and so was auditing against what the appellant documented in their processes.   

Customer complaints

336.       Mr Lee noted that customer complaints were handled by the appellant’s customer relations department rather than by Advantage including claims handling complaints.  As it did not provide claims handling services to other insurers, it referred any complaints on claims handling to the panel insurer.  It reported complaints performance to Advantage and the levels of customer complaints were set out in Advantage’s board report.

Outsourcing by the appellant

337.       Some elements of claims handling were outsourced by the appellant. For example, it did not have its own network of garages so it outsourced that work to partner garages (Nationwide and Fleet Accident Repair Group).  It used solicitors in relation to certain claims but instructed the solicitors individually from a panel (i.e. the solicitors did not have delegated authority).

Claims service from Endsleigh

338.       In the early part of the Period, as noted, Advantage had also received claims handlings services from Endsleigh.  Mr Eagar said that the handling service Endsleigh provided was the same as the appellant’s services but the claims referral threshold was lower (at £20,000 for property damage and £30,000 for personal injury).  The relationship with Endsleigh was relatively new, hence, it was felt that it was inappropriate to give them large limits of delegated authorities at that time.  Any claims over the threshold had to be referred by Endsleigh to the appellant and were managed by the appellant pursuant to its delegated authority.  Once referred, these claims were treated in the same manner as all other claims handled by the appellant (including as regards the thresholds for reporting of large losses). 

Scope of Advantage’s control of claims handling

339.       Mr Lee agreed that Advantage had a close degree of control over the claims-handling procedure both in initially approving the procedure and, when there were changes or amendments suggested by the appellant, in signing them off, as far as they were important ones which could affect Advantage. He thought that was the idea of the service agreement, that Advantage set the high level principles and any major changes should be discussed with them.  But beyond that, on a day-to-day basis, the appellant dealt with tens of thousands of small claims every month, each of which was handled by one of 500 or 600 handlers within the guidelines without any reference to Advantage, other than as regards changes to procedures that were agreed every couple of years.  So the framework was set by Advantage but operationally day-to-day, when somebody called in and said, “I have had an accident”, that was generally dealt with by the appellant without referring back to Advantage.

340.       It was put to Mr Eagar similarly that, as Advantage had the final say over the process manuals, Advantage was providing a framework and controlling the way in which the claims handlers operated.  He said that many smaller claims were really driven by a process which neither Advantage nor the appellant had control of as a large part of that (such as repairs) was driven by the requirements set down by the FCA as regards treating customers fairly.  The handling of other claims, such as whiplash claims, for example, were dictated by the use of the Ministry of Justice portal where all claims were dealt with pretty much in the same way by all insurers. 

341.       Mr Charlton said that Advantage was not involved in non-LLC claims due to the sheer number and volume of them.  In his view, Advantage was certainly not managing the day to day operations.  For those claims, as Mr Eagar noted, Advantage carried out an audit process where it looked at how they were being handled, and looked for things like claims leakage as a result of inefficient claims.  Advantage then reported that back to the managers of the various units within the appellant for the identified areas to be improved. 

342.       He said that the audit of claims demonstrated how efficient the appellant was in claims handling, whether it could improve and whether a particular team had a backlog.  Advantage then dealt with the claims managers at the appellant to try to resolve those situations.  Advantage had no direct control of the appellant in this respect as such, “but obviously it’s very interesting for us because it has a direct impact on our result”.

343.       He also noted that the claims handling guidelines were just that; they were guidelines rather than explicit instructions.  Advantage did not say “open file, turn to page three, do this, do that” at all.  They did, however, set out requirements such as that the claim should be settled within a specified time:  

“So in that sense, we are trying to control the work flow, so it is good, but we don’t have people or any influence over staffing, how the departments are organised, how they achieve what they have to achieve within the guidelines.  That’s up to them.  Do they always achieve the guidelines?  Probably no. But that doesn’t mean we are going to run away from Hastings either.”

344.       He confirmed in re-examination that Advantage had no power to make the appellant employ more claim staff or set the budget for the claim staff or anything along those lines.

Importance of claims handling to the appellant and Advantage

345.       It was put to Mr Lee that claims handling was an important part of the appellant’s business.  He said he thought it was more important to Advantage.  For the underwriter, the capability of the claims-handling and counter-fraud and risk-selection was “massively important”.  It was less important for the appellant in the sense that it made its income from the sale of policies.  But there was a link because the better the underwriter did, the lower the underwriter’s prices would be and potentially the greater the business that would be written by Advantage through the appellant. 

346.       Mr Lee said that claims handling required a constant balance between managing the expenses and managing the indemnity spend on behalf of the underwriter.  So the primary function for the appellant’s claims service was to handle claims efficiently, generally trying to pay as little as possible although acting fairly and morally as regards customers but, as the appellant was an independent business, minimising its expenses.  He thought that there was a “natural play-off” between those factors: you could do as little as possible and have very low expenses but the claims costs for Advantage would then “go through the roof”.  So anything that was potentially risky, such as a change in procedure which might increase litigation in the short term but overall, in the long term, push away fraudulent claims or save the insurer money, was something that should be raised with the insurer. 

347.       Mr Lee said he would generally obtain Advantage’s approval where a claims handling initiative may lead to increased risks for it as insurer (and he gave a number of examples where he had done that).  Advantage approval was also required where a strategic decision or a new initiative might, for example, affect Advantage’s road traffic act liability.  If Advantage decided that a proposal should be implemented, the appellant would ensure that the proposal was built into the appellant’s claims team training and would report on the performance resulting from the proposal at the ORM.

348.       Mr Lee noted that the appellant could, within its delegated authority, change the repair companies and negotiate how much was to be paid for the repair of cars and would only have to inform Advantage that it had made those changes.  However, that was not a strategic change to the claims-handling that could add risk to Advantage’s underwriting portfolio.  Mr Charlton and Mr Eagar both confirmed that Advantage had input into the selection of garage networks, in relation to both large and small claims, because these were large items of expenditure for Advantage

Procedure for dealing with large loss claims

349.       Mr Eagar set out how large loss claims were typically dealt with at LLC meetings as follows:

(1) At least 24 hours prior to the meeting, Mr Lee Noble (the large and complex loss manager at the appellant), or his deputy, sent Advantage large loss reports for the claims to be discussed at the meeting.  Mr Doidge said this included a proposed future action plan relating to a particular claim, which aimed to take the case as far towards settlement as possible and set out a proposed level of reserve and the rationale for the proposed reserve.

(2) Prior to the meeting, Advantage personnel reviewed the proposals made by the appellant and on occasions the underlying files. If they considered the proposal needed to be clarified, they would question the claims handler at the appellant before the case was formally submitted to committee so that they had sufficient information to take decisions at the meeting. This might involve a separate dialogue (such as by way of a telephone conversation) between the appellant’s case handler and Ms McKeever or Mr Eagar.

(3) The appellant’s case handlers presented their files at the meeting, and addressed issues relating to liability and quantum, a recommendation for the right level of reserve (with oversight from Mr Noble) and an action plan. 

(4) Often Advantage and the appellant’s personnel discussed the recommendations at the meeting.  Both Ms McKeever and Mr Eagar had over 25 years of experience in the insurance industry and so had a good understanding of the requirements of the reinsurers which they brought to bear in assessing and discussing the recommendations.

(5) The committee agreed the level of reserve and the proposed future action plan, sometimes on the basis of what was proposed by the claims handler or, as adjusted on the basis of Mr Eagar or Ms McKeever’s review of the file.  The decision as to what reserve to set or whether a claim was to be referred to Advantage’s reinsurers was always taken by the members of the LLC (as Mr Doidge also said).  That was commensurate with the fact that reserves were held by Advantage as insurer and it was Advantage which was responsible for making reports to the reinsurers.

350.       It was put to Mr Eagar that, at the meetings, Advantage was giving instructions to the appellant as to how claims were to be handled.  He said Advantage were “signing off”:  

“Once we have had due consideration and discussion over the matter, we are signing off a plan of action that Hastings had presented to us that will include reserve recommendations, recommendations of how they can resolve liability, for example, how they can ascertain quantum for the case and we are then approving that action plan.  We might throw in a couple of our own suggestions that Hastings may have missed and we have a very experienced - during this period and currently we have got very experienced guys that sit on the committee that can also add a slightly different view of, have you also considered this, have you also considered that, but it is signing off recommendations that they make to us.”

351.       Mr Eagar explained that the document presented by the appellant at the LLC would cover matters such as liability, what investigations they planned to carry out, a review of quantum and then a detailed reserve breakdown in the vast majority of claims that the committee would then go through sometimes line by line.  At the end of it all there would be an action plan that Advantage would either agree or disagree with, and the same for the reserve, and, as experienced claims people, Advantage could add further suggestions to how they could take this claim forward.  But certainly the intention was that this really did take the claim quite a way forward, possibly all the way through to settlement, but certainly there should not be “to-ing and fro-ing” with further referrals to Advantage.

352.       Once the reserve overall strategy was agreed by Advantage the implementation was a matter for case handlers.  For example, if Advantage wanted medical evidence, it would be down to the appellant to approach a medical legal expert to provide that evidence and then report.  Advantage would pick it up as part of the reviews going forward.

353.       Mr Eagar gave two examples of occasions on which, in his experience, the committee made decisions that adopted a different position to the recommendations made by the appellant:

(1) In October 2013, Mr Eagar presented a claim at the LLC in relation to a road traffic accident which left a man with a broken arm, a cut to the head and a minor epilepsy risk. (Unusually he presented the claim as the appellant’s previous large and complex loss manager had just left the business.)  He discussed the case handler’s analysis of the position and the rationale for suggesting a reserve of £100,000, but Mr Godfrey (as managing director of Advantage) disagreed and considered that a £50,000 reserve was more appropriate in the absence of further information.

(2)  In November 2012, there was a case involving a driver who cut across another vehicle on a minor road, leaving a man permanently disabled and unable to return to his former role and full time hours at work. The appellant’s loss manager proposed a settlement offer of £200,000 and an increase in the reserve to £600,000; however the LLC disagreed that the reserve needed to be increased at that stage; Mr Charlton and Mr Eagar decided to maintain the current reserve of £230,000.

354.       Mr Doidge gave a similar description of the process and his understanding of the purpose of the committee accorded with the above description.  Mr Eagar said he or Ms McKeever reviewed every large loss file about every 12 weeks.  He noted that cases could take a long time to be settled.  On average they took approximately three to four years, but claims of catastrophic injuries could take up to six years and child claims up to 10 to 15 years.  Claims settled by way of periodic payment orders for the remainder of the claimant’s life may take up to 50 to 60 years to be finally paid out.

355.       Mr Doidge said that he could tell from the questions that Mr Eagar and Ms McKeever asked at the LLC that they reviewed the relevant case files that were held on the appellant’s IT systems before the meeting.  He thought they would review the accident location and the medical reports and would then debate the respective merits of the reserves the appellant was proposing.

356.       Mr Doidge said that Advantage and the appellant were in regular telephone contact.  For example, Mr Eagar usually spoke to the appellant (usually Mr Noble) on a daily basis to review the reserves on claims or to agree reserve movements between £25,000 (£10,000 during the period) to the large loss threshold of £99,000.  He noted that under the claims handling guidelines, the appellant was required to notify Advantage of these movements, but in practice the appellant sent details of the trigger event for the movement in the reserve, the rationale for the adjustment and an appropriate reserve breakdown and sought approval from Advantage in respect of the change.  These conversations may also include queries such as on where the appellant had got to on the claim, whether an offer had been accepted or what was the practical approach for dealing with the issue.

357.       In Mr Doidge’s experience as an attendee at the LLC meetings, the appellant’s proposed approach to dealing with the claim was discussed and often challenged, in particular, as regards reserves and liability and the apportionment of liability.  He gave the following examples, which were also evidenced in the documents in the bundles such as minutes of the LLC and large loss reports.

(1) In a case referred to the committee in August 2013, the appellant recommended a reserve of £9.5 million as regards a road traffic accident leaving a young woman with paralysis. Following discussion and challenge the committee ultimately authorised a reserve of £8.1 million because contributory negligence was being considered

(2)  In a case involving a catastrophic brain injury to a teenage boy arising from a road traffic collision, the appellant proposed a reserve of an 80/20 liability apportionment (this being a pedestrian claim). The committee approved a reserve based on a 75/25 split in the claimant’s favour as it considered that an 80/20 split was too pessimistic.

(3) In a complicated case referred to the committee in October 2014 the committee requested further information and clarity around the road layout prior to taking a decision on the appropriate reserve for the file. Further investigations at the accident scene were carried out prior to a later referral to the committee in December when the reserve was then approved.

358.       Mr Doidge said that sometimes additional specific agenda items were discussed at the meetings as raised by Advantage.  For example, Advantage raised an issue regarding the accuracy of the opening actuarial reserves that were being set on certain case files, which subsequently led to a review of the approach to this. 

359.       Mr Charlton said that Advantage were heavily involved in the handling of large losses.  They kept separate files in Gibraltar because they were the claims which affected Advantage the most.  Mr Eagar noted that the files contained Advantage’s own notes about the case including details of correspondence with the reinsurers.  Advantage also used its notes to evidence telephone conversations with the appellant.  Advantage then kept under review the indemnity, liability and quantum position for the file.  This file could not be accessed by the appellant as these notes were prepared for Advantage’s own purposes although any operational issues identified were raised with the appellant where Advantage considered that was appropriate.

360.       Mr Charlton’s main involvement with large claims was as regards the reserves position.  Mr Charlton described large loss claims as the most complex and most interesting, for the claims-handler.  He thought that “human nature being human nature”, the majority of large loss claims-handlers, particularly when they were new to the role, “love to put a large loss on at a large level” of reserve, so when it came to settlement they appeared to have been very successful. He noted that the fact that the initial figure was wrong in the first place did not seem “to go into their psyche”.  So this was an issue he was very alive to when considering large loss reserves, on the LLC, to listen to the argument and see whether the handlers were firm on the information they had on the best estimate.  He said that, equally, it was not good to be too imprudent, because that lead to under-pricing and a bad result. So it was a balancing exercise and that is where he personally inputted on the LLC, essentially, where he considered there was insufficient evidence for a reserving position and the claims-handler has been “a bit gung-ho with it”.  

361.       There was also evidence from the documents in the bundles that Mr Charlton was actively involved in the discussions at the LLC meetings and evidence of the on-going dialogue between Mr Eagar and the appellant on large loss claims.  For example the minutes of the LLC on 5 October 2009 record an employee of the appellant provided advice but Mr Charlton requested that particular steps be taken (namely, that surveillance be undertaken in line with a recommendation that had been made by solicitors).  In the minutes of the meeting on 14 December 2009 it was recorded that Mr Charlton asked for a claim to be “reported to reinsurers along with details of our current reserve, which is within our reinsurance retention”.  The minutes of the meeting on 25 February 2010 record that:  “As discussed previously between GE (Gary Eagar of Advantage) and MD (Mark Danby of Hastings), we are able to review reserve six months post-accident.”

362.       Mr Eagar confirmed the following as regards the process for settling claims:

(1) As set out above reserves for large losses were set at the LLC.  (As set out above, reserves for claims under the threshold were set by the appellant and it settled the claim under its delegated authority, subject to.  Advantage’s audit programme). 

(2) If the settlement offer (including costs) was less than the reserve or within £25,000 of the reserve and Advantage had agreed the reserve, the appellant had total discretion to settle the claim.  Ultimately, Advantage expected the appellant to try to settle all cases as economically as possible.  Advantage audited claims to monitor whether the appellant was meeting expectations.  The appellant had to weigh up the pros and cons of becoming embroiled in settlement negotiations with a view to trying to reduce the cost of the settlement, as this potentially increased the costs incurred by legal representatives of the injured party.

(3) If the settlement offer was higher than the reserve by more than £25,000, the appellant obtained specific approval from Advantage in order to amend the reserve to be able to settle at the higher amount.  Mr Eagar usually attended settlement meetings relating to the very largest claims, those valued at £3 to £4 million upwards, where Advantage was sharing the risk with reinsurance partners and any cases where it was at risk of the claim settling as a periodic payment order.  He was particularly anxious to attend when the reinsurers attended the meetings (for claims which are large enough to be covered by excess of loss reinsurance) so that he could help to guide the reinsurers through the process and manage their expectations through to settlement of the case.

(4) For larger claims and claims that had been reported to Advantage’s reinsurers, Advantage had to provide the appellant with an express communication setting out the final settlement figure if that amount was over £250,000. This would be done by a telephone conversation between the appellant’s claims handler and Ms McKeever or Mr Eagar that was recorded in notes on the appellant’s and Advantage’s systems, or by email exchange granting the appellant specific authority to settle the claim up to a defined amount. The appellant was required to notify Advantage in advance of any payments being made in settlement of claims in excess of £250,000.  Mr Doidge said that, in practice, the appellant asked Advantage to give its express approval to cases where it proposed to offer to settle a claim for an amount in excess of £250,000. The appellant was also required to make certain additional notifications to Advantage in respect of claims with a total incurred in excess of £375,000 or other claims which were required to be reported to reinsurers.

(5) Occasionally it was necessary to instruct lawyers to conduct litigation or other dispute resolution processes.  The appellant could appoint the lawyers without needing to seek authorisation from Advantage, pursuant to the claims handling guidelines and the services agreement.

363.       Mr Eagar noted that the appellant had a panel of law firms which it used.  He was involved in the selection process by which these firms were selected.  He was also on the committee which was in the process of reviewing the garage network providing mechanical repairs, which represented a significant proportion of the claims handling spend. He confirmed that all of the customer-facing, all of the dealings with the claims with solicitors, the customer and third parties, was done by the appellant.

364.       It was noted to Mr Eagar that, although the appellant actually appointed the lawyers, under the agreement Advantage could object to the appellant appointing particular lawyers.  He said if there were concerns around their performance or their conduct Advantage would intervene but Advantage would not just turn around and say it did not like a particular firm.  There was a tender process.  There were two very large firms on the panel that were both competent.  Advantage had no objections to appointing them on any claims.  That is why they were selected because they were very good. He could not see any occasion for Advantage preferring one firm over another.

365.       Mr Charlton said that his understanding was that “lawyers were appointed by Hastings in the normal course of handling a claim.  When it came to a large claim, you would look for a specialist lawyer to handle that large claim”.

 

 

Relevance of claims handling to reinsurance

366.       Advantage was responsible for reporting claims which it was required to report to the reinsurers.  The appellant had to report all such claims to Advantage to enable it to do so.  Advantage sent the appellant a copy of the reporting forms that Advantage sent to its reinsurance brokers so that the appellant was aware how much of the relevant file was seen by the reinsurers.  This meant that review visits by the reinsurers to the appellant’s premises ran more smoothly.  It was easier simply to copy in the appellant’s  case handler to all reinsurance correspondence for the handler to record on his file which the reinsurers could then review as required (as they rarely brought their own files with them during the reviews). 

367.       Reinsurers were concerned to know how claims were handled.  Therefore, they asked for review meetings which until recently mostly took place at the appellant’s premises in Bexhill.  Mr Eagar held review meetings with the reinsurers on average 10 to 12 times a year.  He usually tried to attend the visit by the reinsurers in the UK, or he dialled in by telephone from Gibraltar, in order to make sure that Advantage’s reinsurers were content with the claims handling and other intermediary and underwriting support services and that they did not have concerns relating to the handling of a particular claim, incorrect reserves or process issues.

368.       The frequency of meetings with reinsurers has increased over the years with the growth in Advantage’s book.  However, as the Gibraltar market has grown (so there are several Gibraltar insurance companies that a reinsurer may wish to review in a visit) and the book has developed to contain more large loss claims, reinsurers increasingly wished to visit Advantage’s offices to review first-hand its systems and procedures for managing the appellant’s claims handling services.

369.       Mr Doidge and Mr Noble sometimes attended these meetings to provide Advantage with additional support.  They generally attended meetings held in Bexhill, and occasionally meetings held either in London, at the reinsurance broker, or in Gibraltar.  This was because the appellant’s claims handling employees managed about 50 to 90 case files each.  By contrast, Ms McKeever and Mr Eagar managed about 200 files each.  Therefore, the appellant’s team members had greater knowledge of the detail of their cases.  This did not affect Advantage’s involvement in decision making at LLC meetings, as Advantage was afforded the opportunity of reviewing a file prior to making the decision. This was much more difficult in meetings with reinsurers who reviewed 20 or so cases in a day and were seeking immediate answers to their questions.

370.       Although the reinsurers could access the appellant’s files in Gibraltar, by using Advantage’s IT system with permission from Advantage, the purpose of holding some meetings in Bexhill was to allow the reinsurers to examine the non-LLC claims handling work, that being the bulk of the work that was carried out by the appellant. Although this information was available on Advantage’s IT system, reinsurers preferred to discuss the processes, rather than review a computer file.

371.       It was put to Mr Eagar that the appellant’s staff attended meetings with the reinsurers as there was a very close relationship between the appellant and Advantage on an integrated model of insurance.  Mr Eagar said Advantage was responsible for reporting all the claims to reinsurers.  They were its reinsurers, not those of the appellant, as the intermediary.  In the same way reinsurers would also meet the panel solicitors who would also provide assistance.  So he did not see it that way.  He saw it as part of the appellant’s claims handling function and its team’s closer knowledge of the files.  It was for those reasons that the appellant supported Advantage’s meetings with its reinsurers.

Auditing of claims handling

372.       Both the appellant and Advantage carried out auditing work on a monthly basis. Advantage mainly focused on the technical areas, such as personal injury and fraud, looking at areas such as indemnity, liability and quantum decisions. The appellant focused on the customer areas such as customer care, total loss and recovery.  There was some overlap between the auditing work they each did, for example, in relation to leakage.

373.       The large loss team at Advantage had a close working relationship with the internal audit team at the appellant.  Advantage’s audits allowed it to maintain a wide view over the appellant’s work.  On average Advantage conducted 150 “technical audits” per month.  It operated a risk based audit approach which was intended to ensure that it audited the business in accordance with the risk within each area.  For example, approximately 40% of claims payments were incurred on personal injury claims, therefore, approximately 40% of audits focussed on those claims. The remaining 60% was split equally between third party property damage and owner damage, which reflected the approximate split in claims reserved between the two areas.  The audits were widespread and followed a schedule throughout the course of the year.  The auditors at Advantage selected their audit sample in accordance with the nature of the audit that was being undertaken and following the criteria stipulated within the audit schedule. The file was then audited against a detailed pre-defined audit question set, looking at issues such as indemnity, liability, quantum, reserving and financial leakage.

374.       Mr Eagar said that, of the 72,000 or so non-LLC claims per year, roughly about 800 a month were audited so around 10,000 in a year.  He said that there were different levels of audits; some would look at individual payments and others at the handling of the claim.  An audit may well only look at a specific item rather than the whole claim.  The appellant simply got on with around 80 to 85% of claims but Advantage could obviously monitor their performance through the management information to see what levels of settlement they were achieving on particular types of claims.

375.       Advantage wanted to understand how, and why, the appellant changed its processes in order to react when the insurance, legal or consumer market evolved. Advantage also examined how the appellant resolved issues and whether there was a weekly improvement in relation to any identified issues. Therefore, it was important to keep a dialogue with the appellant open, in particular through the weekly and monthly auditing undertaken by Advantage and the audits performed by the appellant which Advantage reviewed as part of the management information and by discussing the issues that were identified through these audits with the appellant.  The main contact for this purpose was the audit manager at the appellant but in addition issues could be raised at the monthly ORM.  Mr Eagar said that regular communication between Advantage and the appellant was the best way of ensuring that issues that were identified through Advantage’s audits were resolved in a way that satisfied Advantage as being appropriate.

376.        Mr Eagar had oversight of the audit work Advantage carried out in relation to the appellant and he presented management information about the audits.  His report to the Advantage board included a very high level overview of the audits, and he commented on claim trends, such as, amounts of new claims, amounts that were paid in claims and provided a summary of all losses over £500,000.

377.       Around 8 to 10 times each year he spent a day face-to-face with the appellant’s technical claims director, or one of his senior employees who directly reported to him, such as Mr Doidge.  Roughly half of these meetings took place in Bexhill and the rest in Advantage’s offices in Gibraltar.  He also spoke via telephone with senior claims management at least fortnightly.  This involved reviewing issues including trends at the appellant, market trends, audit issues, resourcing issues, supplier issues and legal or regulatory charges.

378.       The Advantage employee conducting the audit produced a report in relation to each claim that was audited. The top sheet of the report summarised the themes identified by the audit (for example, the fact that indemnity queries have not been checked) and identified all the audited cases where a particular recurring issue had occurred.  Problematic cases were highlighted.  For example, Advantage might notice a lot of issues arising from a particular claim handler at the appellant and would alert it to this trend.

379.        Issues were raised within the audit report which was sent directly to the appellant. The appellant’s audit manager was responsible for collating the appellant’s management responses to Advantage’s audits (the appellant had one month in which to add their management comments to the report) and sending the report back to Advantage.  Advantage then decided whether or not it agreed with comments and whether to amend the report. 

380.       Mr Eagar said the vast majority of the issues the audits revealed were down to handler error.  It was a busy department with a lot of work.  Nothing the team saw in audits was out of line with what would occur in the majority of motor insurers.  He was asked if Advantage would suggest an actual change itself if there was, for example, a persistent handler error which may indicate that there was something wrong with the way the computer was set up.  He said if the problem was an individual, the audit would name that person and it would then be for the appellant’s management to deal with that individual as they saw fit.  It was not for Advantage to dictate that.  He could not think of any instance where the team had seen an issue with any process of the appellant’s where Advantage had said it had to be changed.  A lot of the issues the team saw were just down to the usual backlogs of work because it was a busy claims department.

381.       The audit report was not sent to Advantage’s reinsurers unless they asked for it. The Advantage claims auditor’s feedback was also circulated within Advantage. Advantage had 4 or 5 board meetings every year at which the audit reports relating to the appellant were discussed.  Prior to the board meeting, Mr Eagar prepared a summary of all the audit reports and of the percentage of problematic cases over the period.  He then presented the six most recurring problematic themes to the board, which were included in the board report.

382.       Mr Eagar noted that the audit process was broadly similar to the process followed when Advantage used Endsleigh to handle some claims. However, as Advantage did not have direct access to Endsleigh’s systems, Advantage employees had to physically attend their offices in Cheltenham to review files.

383.       Mr Doidge’s evidence accorded with the above.  He noted that under the appellant’s delegated authority the technical claims team handled and settled most claims up to the threshold but Advantage was aware of how the team handled the claims through its monthly audits of such claims.  He also noted that Advantage also audited leakage in relation to personal injury losses and fraud.  Advantage’s audits assessed the quality of service being provided by the appellant and revealed cases where the appellant had made a mistake in handling a claim or was slow at paying compensation.  He noted that the appellant’s claims audit manager was required to respond to Advantage’s audit report within one month and a procedure to fix the issue (such as additional training for claims handlers) would be implemented by the manager in consultation with Mr Doidge.

384.        Mr Doidge confirmed that he visited Gibraltar periodically (approximately 2 to 3 times per year) to give Mr Godfrey a business update, which included giving him an overview of the claims handling business and of what initiatives the appellant was carrying out, such as, recruitment drives.

385.       He said that the appellant’s work in relation to claims handling and reserving decisions was also checked by Advantage’s reinsurers.  Advantage was the key point of contact in relation to reinsurers and Advantage organised the meetings with the reinsurers, including the meetings at the appellant’s offices during which the reinsurers review the appellant’s files and the committee process. 

386.       Further, an audit team at the appellant carried out regular leakage audits on a sample of claims. The appellant’s audit team now audit all areas relating to the appellant’s claims handling services except fraud, which is dealt with by the Insight (counter fraud) team leaders.

387.       Mr Doidge noted that there was some duplication involved in both the appellant and Advantage auditing claims, but this was natural because the appellant wanted to ensure that it was providing a suitable service and was complying with the claims handling guidelines.  His understanding based on conversations with Mr Eagar was that Advantage wanted to monitor the quality of service provided and had to oversee the appellant’s work to comply with regulation in Gibraltar.

Fraud initiative

388.       Mr Lee said that a counter fraud team was created by the appellant about three years ago by bringing together in a single team those involved in fraud issues in the claims, underwriting and broking teams.  Mr Lee considered that this would allow the appellant to introduce innovations in the service provided which would help Advantage and other insurers improve their loss ratios.

389.       The first idea Mr Lee developed was software called “Quote Manipulation Software”.  The appellant could see from the aggregate markets that the amount of misrepresentation of risk was increasing with the effect that insurers could potentially lose money because of the level of quote manipulation and misrepresentation.  Using the software the appellant took all the quotes that customers were given over a period of about 10 days from aggregator websites and analysed them to identify instances in which the information provided by a customer was inconsistent (for example, in relation to any speeding convictions that they might have) which might indicate that the customer was manipulating their details to obtain a lower premium.

390.       When developing this initiative, the appellant modelled the effect it would have on Advantage’s loss ratio and then approached Advantage to present the analysis and request a financial contribution from Advantage as it would benefit from the initiative if successful.  The proposal was that Advantage would reduce its net premium rates by an amount that would share the benefit of the expected improvement to its loss ratio.  The appellant as intermediary would obtain its share of the benefit through the opportunities afforded by net premium rates that were more competitive in the market.  For Advantage to be able to make an upfront reduction to the net premium rates they needed to be confident that the analysis and projections for the benefit of the initiative to their loss ratio were robust, otherwise they would have waited to see the improvement in the loss ratios as they matured over the next couple of years.  Mr Lee explained that the appellant then offered that quote manipulation protection to other underwriters in exactly the same way, asking for a cut in rates in return. 

391.       Counter fraud initiatives were discussed and monitored at monthly Insight committee meetings, which were attended by Mr Godfrey, Mr Eagar and Ms McKeever from Advantage, the appellant’s counter fraud director, Mr Hoffman and representatives from the appellant’s IT, claims, compliance and retail departments.   The meetings involved both decision-making and also discussion of a monthly management information pack.  The pack set out what percentage of cases had been through the fraud department and measures of the benefits realised through the counter fraud strategy.

392.       Other specific issues, such as operations to counter organised frauds (for example ghost brokers) and fraud cases which were being taken to trial were reported by the appellant to Advantage and also discussed at these meetings.  Advantage was primarily concerned with ensuring that so far as possible its loss ratio was minimised. Advantage was involved in decisions which affected reputational or litigation risk. These matters may also be considered at LLC meetings.

ORMs

393.       Mr Godfrey described these meetings as taking place to review the key performance indicators for the delivery of the appellant’s services and Advantage’s financial performance. Mr Lee confirmed that he and his counterpart in the appellant’s underwriting team reported to Advantage at these meetings on the underwriting performance and counter fraud, telematics and claims handling.  He described the meetings as mainly focused on the appellant’s operational performance, for example, whether it had enough staff, what changes were proposed or what changes it was making with processes and procedures within its departments.

394.       Mr Lee said that, prior to the meeting, a standard information pack was provided to the attendees.  This included an executive summary of the activity in the claims department, recruitment issues, head count trackers and the appellant’s profit and loss account.  Mr Godfrey noted that there were underwriting reports and a finance report as well as “headcount and performance scorecards” for the appellant’s bulk claims handling department.  In relation to underwriting support services, the report in the pack included a summary of information on frequency of losses (the number of claims as a percentage of the policies in force) and the loss ratio. 

395.       Mr Godfrey said that the main reports discussed at these meetings were (a) the operational “Insurer Services Scorecard”, that included detailed information on operational key performance indicators for claims operations, claims fraud, underwriting operations and account performance, (b) an underwriting report that addressed retention rates, premium volumes and market analysis, summary of product updates and an audit summary data and (c) a claims report, that covered areas such as the volume of claims, backlogs, complaints and average claims values.

396.       Mr Lee said that the profit and loss account was included, in particular, because Advantage examined the appellant’s salary and other expenses: it was important for Advantage to be aware of this because, for example, if the appellant was underspending against budget as regards salary (in an effort to boost profits), it was an indication that the appellant was understaffed and that its service to Advantage and, ultimately, Advantage’s loss ratio, was going to suffer.  Advantage also reviewed the appellant’s claims income which comprised commissions from third parties in respect of claims handling activities for non-fault policyholders such as referring them to providers of credit hire vehicles.  This was relevant as the income had to be taken into account in determining the appropriate arm’s length payment to be made by Advantage to the appellant for claims handling services.

397.       Mr Lee noted that an insurance company which outsourced claims handling had to make sure that the service provider actually had the capacity to service claims.  Backlogs in claims handling were likely to have a severe, negative effect on the insurer’s loss ratio and, therefore, its profitability.  The danger was that the problem would not be picked up through changes to reserves on files and it would only be 3 or 4 years down the line, when the claims had fully developed, that the true cost would be apparent.  Mr Lee had experience of this occurring in his previous employment. 

398.        In relation to claims handling, the report in the pack included analysis relating to the average severity (and monthly cost) of claims by reference to the “three perils” or categories of loss, being accidental damage, third party property damage and bodily injury, as well as discussion of trends, call centre response times and other service level indicators.  Given the volume of claims (around 12,000 a month), unless there were unusually contentious claims, individual claims were not discussed because, except as regards large loss claims, they were handled by the appellant under its delegated authority as set out above.

399.       Mr Lee said that if Advantage wished to raise an issue or noticed a problem in the reports, they raised it with the appellant at the meeting or prior to it and the appellant then proposed a plan for resolving the issue.  Advantage then decided whether they were satisfied with the proposed plan and, if they were not, a revised course of action was generally agreed.

400.       Mr Godfrey said that the discussions on the claims handling process centred on trends, average claims settlement figures, call centre response times and other service level issues.  Advantage questioned and challenged the appellant, for example, where Advantage spotted a trend in the claims handling area.  He said the appellant’s team would examine the issue and then report back with suggestions for potential changes to claims handling procedure. The appellant’s team routinely sought Advantage’s approval in relation to any potential changes to those procedures

401.        Mr Lee noted that if the issue or problem caused the appellant not to meet targets that were agreed by the parties, such as targets relating to the level of leakage, the appellant’s performance commission was reduced in accordance with the current terms of the services agreement.  He said that the final decision as to what constituted leakage was taken by Advantage, who could take a more independent view than the appellant on that.  Advantage sent its findings on leakage (in the form of case by case reviews) to the appellant’s claims audit manager.  The appellant had an opportunity to make submissions in relation to these findings and then Advantage produced leakage analysis which was included in financial reports (including the appellant’s monthly audit report). 

402.       Mr Godfrey said that the areas of contention between the appellant and Advantage as regards small loss claims often centred around leakage as there was an element of subjectivity as to what should be classified as leakage.  He said that whilst these matters were discussed, Advantage had the final say on leakage as it prepared the leakage audit reports. 

Discussion - overview

403.       The question for the tribunal was whether the appellant’s supplies of broking, claims handling and underwriting services were made by the appellant to Advantage at its BE in Gibraltar or, at a FE in the UK. It is only if the supplies of services were made to Advantage at its BE in Gibraltar, that the appellant can recover or obtain credit for its input tax incurred in making those supplies. 

404.       To re-cap, the dispute between the parties was as follows.

(1) HMRC argued that Advantage had a UK FE, comprising the appellant’s human and technical resources, through which it made supplies of insurance to UK customers.

(2) The appellant disputed that Advantage had any such UK FE.  In its view supplies of insurance were made from Advantage’s BE in Gibraltar.

(3) HMRC asserted that it followed from the first point that the supplies of services were made by the appellant to the UK FE for use at the FE in making the supplies of insurance.  As noted the appellant disputed that there was any such FE but argued that, on that analysis, in any event many of the services (such as the underwriting support and claims handling as regards large losses) were received and used at the BE in Gibraltar. 

(4) HMRC said that whilst primacy is usually accorded to the BE, in their view, this would lead to an irrational result such that the FE should be regarded as the place of supply/belonging. The appellant said that using the BE would not in fact give an irrational result.

405.       On that basis it is necessary to consider first whether Advantage had a FE in the UK both as regards (a) supplies of insurance it made and (b) as the recipient of the supplies of services made by the appellant.  If it did have a FE in the UK, the further question is whether Advantage’s BE or FE is to be preferred as the place of supply/belonging. We have, therefore, first considered the case law on when there is a FE and when it is appropriate to have regard to that establishment rather than the BE. 

406.       We note that, whilst the case law pre-dates the definition of FE in the Regulation, it has been held by the CJEU to be relevant to both the position before the introduction of the Regulation and that applicable from then on in the case C-60/125 Welmory sp. z o.o. v Dyrektor Izby Skarbowej w Gdańsku [2015] STC 515 (see [457] to [482]).  The discussion in the older cases relates to the place of supply rules in article 9 of the Sixth Council Directive 77/388/EEC of 17 May 1977 (“article 9”) which were substantially the same as those applicable in the Directive before 1 January 2010 (see [29] above).

Discussion - Case law

Bergholz and Faarborg-Gelting

407.       In Case C-168/84 Gunter Berkholz v Finanzamt Hamburg-Mitte-Altstadt [1985] ECR 2251, the issue was whether an undertaking based in Hamburg, which operated gaming machines on board two ferries, had a FE on board the ferries as a result of the presence of the machines.  As the facts were set out by the Advocate General, the only resources present on the ferries, other than the machines themselves, were two persons who were employed for two hours a week to keep the machines in good condition and repair and who collected the money from them.  The CJEU held (at [18]) that there was no FE and in doing so set out guidance on what constitutes a FE, which has been cited extensively in the subsequent cases:

“It appears from the context of the concepts employed in Article 9 and from its aim ….that services cannot be deemed to be supplied at an establishment other than the place where the supplier has established his business unless that establishment is of a certain minimum size and both the human and technical resources necessary for the provision of the services are permanently present.  It does not appear that the installation on board a sea-going vessel of gaming machines, which are maintained intermittently, is capable of constituting such an establishment, especially if tax may appropriately be charged at the place where the operator of the machines has his permanent business establishment.”  (emphasis added)

408.       The CJEU also said, at [17], that in determining the place of supply, reference should be made to a FE only if the reference to the place where the BE is located does not provide a rational result:

“According to Article 9(1), the place where the supplier has established his business is a primary point of reference inasmuch as regard is to be had to another establishment from which the services are supplied only if the reference to the place where the supplier has established his business does not lead to a rational result for tax purposes or creates a conflict with another Member State.”

409.       The Advocate General had taken the contrary view that the principle should be followed that tax was due at the place of consumption.

410.       This approach was followed (although there is little reasoning in the decision) in Case C-231/94 Faaborg-Gelting Linien A/S v Finanzamt Flensburg [1996] ECR I-2395, [1996] STC 774.  In that case it was held that a company established in Denmark, which operated a ferry service between Denmark and Germany, did not have a FE on the ferry as a result of the on-board restaurant service provided for passengers.  The CJEU referred, at [16] and [17], to the approach in Bergholz and concluded, at [18], that the FE test in article 9(1) did not seem to apply “to a place supplying restaurant services on a ship, especially where, as in this case, the permanent establishment of the operator of the ship affords an appropriate point of reference”.  They concluded at [19] that restaurant transactions are to be regarded as supplies of services which are deemed to be carried out at the place where the supplier has established his business.

 ARO Lease

411.       In Case C190/95 ARO Lease BV v Inspecteur van de Belastingdienst Grote Ondernemingen te Amsterdam [1997] STC 1272; [1997] ECR I-4383 a company established in the Netherlands, leased cars to customers in Belgium acting through self-employed intermediaries established in Belgium, who received commission for their services.  The company did not have office premises or storage premises for the fleet of cars in Belgium.  The issue was whether the company made supplies of the leasing of cars in Belgium or the Netherlands. 

412.       At [14] the court noted that it had been held in the CJEU that “since forms of transport may easily cross frontiers, it is difficult, if not impossible, to determine the place of their utilization and that in each case a practical criterion must therefore be laid down for charging VAT”.  Therefore, the Sixth Directive provided that the hiring out of all forms of transport should be deemed to be supplied “not at the place where the goods hired out are used but, with a view to simplification and in conformity with the general rule, at the place where the supplier has established his business” (citing Case 51/88 Hamann v Finanzamt Hamburg-Eimsbüttel [1989] ECR 767, at [17] and [18]).

413.       The court referred, at [15], to the decision in Bergholz and endorsed the view that the BE is the primary point of reference and couched the approach to determining whether there is a FE in similar terms to those used in Bergholz:

“It is clear from the aim of Article 9 and from the context in which the concepts are employed that services cannot be deemed to be supplied at an establishment other than the main place of business unless that establishment has a minimum degree of stability derived from the permanent presence of both the human and technical resources necessary for the provision of the services..”. (emphasis added)

414.        The court continued, at [17], that consequently, in order to be treated, by way of derogation from the “primary criterion” of the main place of business, as the place where services are provided:

“an establishment must possess a sufficient degree of permanence and a structure adequate, in terms of human and technical resources, to supply the services in question on an independent basis.” (emphasis added)

415.       In considering whether the facts were sufficient for the company to be regarded as having a FE in Belgium the court noted, at [18], that the services supplied in the leasing of vehicles:

“consist principally in negotiating, drawing up, signing and administering the relevant agreements and in making the vehicles concerned, which remain the property of the leasing company, physically available to customers”.

416.       The court concluded at [19] that:

“when a leasing company does not possess in a Member State either its own staff or a structure which has a sufficient degree of permanence to provide a framework in which agreements may be drawn up or management decisions taken and thus to enable the services in question to be supplied on an independent basis, it cannot be regarded as having a [FE] in that State”.

417.       They considered it clear, at [20], from the wording and aim of article 9 and from the judgment in Hamann cited above, that neither the physical placing of vehicles at customers’ disposal nor the place at which they are used could be regarded as a “clear, simple and practical criterion, in accordance with the spirit of the Sixth Directive, on which to base the existence of a [FE]”.  Also, at [21], the fact that customers chose their vehicles themselves from Belgian dealers had no bearing.  Nor could the self-employed intermediaries “who bring interested customers into contact with [ARO Lease]” be regarded as permanent human resources within the meaning of the case-law.  The fact that the vehicles were registered in Belgium, where road tax was also payable, “relates to the place where they are used, and that factor, in accordance with the case-law cited above, is irrelevant”.  At [22] they held, therefore, that on such facts the services could not be regarded as provided from a FE in Belgium.

418.       At [23], they noted that submissions were made that account must be taken of “economic reality” when applying article 9 to forms of transport, and the place where the services were provided must be held to be the place where the business in question is actually carried on.  They held, at [26], that that interpretation would “run counter to the intention of the legislature, which, taking economic reality into account, as regards forms of transport, has decided to introduce a clear, simple and practical criterion, namely, the main place of business or that of a [FE]”.  

DFDS

419.       Case C-260/95 Commissioners of Customs and Excise v DFDS A/S [1997] STC 384, [1997] ECR I-1005 concerned the place of supply of package tour services to customers in the UK and Ireland made by a Danish company, DFDS, through the agency of its wholly owned English subsidiary.  It was not disputed that DFDS was liable to account for VAT under the special rules for travel agents and tour operators in the Tour Operators Margin Scheme (established by article 26 of the Sixth Directive, now article 307 of the VAT Directive) (“TOMS”). 

420.       In brief, that scheme contains special VAT rules for travel agents and tour operators essentially where the agent acts in its own name and not as an intermediary. It is designed to avoid practical difficulties for such agents in accounting for VAT as they typically provide services to customers in a number of countries.  In this case it was DFDS which, as tour operator, fell with the scope of the scheme and not the English subsidiary, which acted as an intermediary. 

421.       The place where such services are taxable is determined as the member state in which the travel agent has established his business or has a FE from which it provides the services (under article 26).  The UK government took the view that the services were supplied by DFDS through a FE in the UK in the form of the English subsidiary.  DFDS argued that the services were supplied in Denmark where it had its BE.

422.        The Advocate General set out, at [3] and [4], the relevant facts.  In summary:

(1)  The parties concluded an agency agreement to govern relations between them whereby the English company was appointed general sales and port agent for DFDS and was entrusted with making reservations (throughout the UK and Ireland) for the passenger services operated by DFDS. 

(2) In addition the English company was required to carry out a number of tasks which included: providing assistance to DFDS in supervising and controlling tours; making available qualified sales and operational personnel; consulting DFDS on the employment of management staff; obtaining the approval of DFDS before concluding any major contracts and for the appointment of advertising and public relations agents.

(3) The English company was also required to promote its commercial image in accordance with DFDS’ strategies and within the financial constraints specified by it.

(4) The English company had to deal with passengers’ complaints and was subject to other obligations in accordance with the company’s policy, including refraining from taking any legal proceedings without DFDS’ prior approval.

(5) The English company was not authorized to work for other passenger transport companies without DFDS’ prior consent.

(6) In return for such activities DFDS paid a gross commission of 19% on all fares sold by the English company.

(7) When called on to do so (either directly by a customer or through a travel agency) the English company had access to DFDS’ central computer in Copenhagen, which contained information on the availability of passenger space and hotel accommodation.  Where the trip or accommodation requested was available, the reservation was accepted and the English company provided the passenger with the requisite documentation.  That documentation was issued in the name and on behalf of the Danish company.

(8) The discretion enjoyed by the English company in matters of pricing was extremely limited.  It had to observe the framework laid down by DFDS in consultation with the English company. At the end of each month, the receipts of the English company were transferred, after deduction of the agreed margin of 19%, to DFDS’ account.

(9) The English company thus carried on directly the business of marketing and advertising, but coordinated its activities with the commercial division of DFDS.

423.       The Advocate General noted, at [11] to [13], that the special TOMS rules were introduced to avoid the practical difficulties travel agents and tour operators would otherwise face in accounting for tax.  However, the place of supply rule in article 26(2) was similar to that in article 9 such that it was necessary to consider how that provision was construed by the courts.

424.       Having referred, at [17] and [18], to the test in Bergholz, at [19], he said it was helpful to consider decisions in competition matters on the concept of agency.  He noted, at [20], that those cases concerned whether the agent is “independent” of the principal.  He referred, in particular, at [21], to a case where it was held that a travel agent was independent where “he sells travel organised by a large number of different tour operators and a tour operator sells travel through a very large number of agents” and that such a travel agent “cannot be treated as an auxiliary organ forming an integral part of a tour operator’s undertaking.”  At [22] he concluded from the criteria set out in the cases that the company could not be regarded as “an independent agency” for reasons relating both “to the structure of its ownership and functional aspects”:

“in the first place, the ownership of all the capital of the subsidiary company is indicative of its 'dependency' on its parent; and secondly…….the English company….does not market tours organized by a very large number of tour operators.  Rather, its contractual link with its parent means that its agency business can be carried on only in relation to the parent, unless the latter has expressly consented otherwise.  Besides, as the agency agreement defines the relations between the parent company and the subsidiary, the latter has no effective independence from the former in the conduct of its business.  The same conclusion follows from a number of points…… in particular, the need for prior approval from the parent company regarding management of the subsidiary company, such as the appointment of senior staff…, the conclusion of major contracts, the appointment of advertising and public relations agents……, and the lack of any discretion in setting the prices of services.  All in all, it seems to me that, having regard to its legal form, the English company acts as an auxiliary to the parent company”.

425.       He continued, at [23], that if, as had been held in another case, the criterion of risk was relevant, his conclusion was the same.  The English subsidiary “does not seem in fact to bear any financial risk under the contracts it concludes with consumers in the course of its agency work” on behalf of DFDS such that it was, he said at [24], “an auxiliary organ forming part of the Danish company from the economic point of view”. 

426.       He noted that it remained to be considered if it was a FE of the Danish parent.  In that context he said, at [25], that he regarded as of importance the cases referred to by the UK government of Case C-221/89 Factortame [1991] ECR I-3905 and Case 205/84 Commission v Federal Republic of Germany [1986] ECR 3755 as regards the meaning of establishment.  He said that Factortame makes it clear that the concept of establishment involves “the actual pursuit of an economic activity through a FE in another member state for an indefinite period”.  He noted that the other case held that for Treaty purposes an undertaking could have a permanent presence in another member state even if there was no branch or agency, but it “consists merely of an office managed by the undertaking’s own staff or by a person who is independent but authorized to act on a permanent basis for the undertaking, as would be the case with an agency”.

427.       At [26] he said that in his view the requirements laid down by the court in those cases were satisfied:

“There is actual pursuit of an economic activity, it is pursued for an indefinite period and there is a [FE].  All those points are confirmed by the detailed examination of the facts…..the most important of which…..is the fact that the English company has about 100 employees…..in addition, the service offered to consumers originates in the United Kingdom.  The contract is concluded in the United Kingdom; it may be presumed that payment is made in local currency; any complaints from customers will be dealt with by the English company; and the parent company reimburses any expenses incurred by the English company in legal proceedings to protect its interests”.

428.       At [27] he concluded that on the basis of the above the English company “fulfils the conditions for classification as an establishment”, as defined in Berkholz:

“There is 'permanent presence of both the human and technical resources necessary for the provision of those services'.  There is everything necessary for a [FE].”

429.       He considered that the circumstances in DFDS were entirely dissimilar to those in ARO Lease.  In particular, the Netherlands company in that case simply had no place of business in Belgium.

430.       At [28] he turned to whether the supplies of services should be regarded as made from Denmark or the UK noting that the FE is to be taken into account only in the alternative, as set out in Bergholz.  He noted, at [30], that the UK government preferred the approach of the Advocate General in Bergholz that regard must be had to the place of consumption.  At [32] and [33], he said he aligned himself with that view as he saw two difficulties with using the BE as the place of taxation. 

(1) That did not accord with the place at which the service was supplied to the consumer, that being “the basic criterion: the VAT system must be applied in a manner as far as possible in harmony with the actual economic situation” such that the FE must not automatically be regarded as subordinate to the BE. 

(2) In addition, using the location of the BE would lead to potential “distortion of freedom of competition and other, more wide-ranging repercussions for the business world”.

431.       He continued that, as the provision itself is silent on which criterion is to prevail and the preamble to the directive does not provide any assistance, he would rely on the “general principle that [VAT] should be charged at the place of consumption and hence give preference to the criterion which enables the supply of services to be located more accurately”.  In his view there was “no doubt that the more appropriate of the two for that purpose is the criterion of the FE, which is clearly more precise”.  He considered that to accept the criterion of the registered office in such a case would result in “distortion of competition between undertakings operating in the same market” and “tour operators in the UK would be discriminated against for establishing their headquarters in one place rather than another”. 

432.       He made a number of other points in support of his view (at [34]) :

(1) It was necessary to consider the wording and purposes of article 26(2) noting that it expressly lays down two criteria rather than just one; the criterion of where the service emanates is no less important than the criterion of where the supplier has its BE. 

(2)  He said that in reading the judgment in Berkholz, “due account must be taken of the facts of that case”. The scope of that judgment “must not be unjustifiably extended by construing it as meaning that the criterion of the establishment from which the services are provided is necessarily merely residual”.

(3) Recognising that the place of the service is the location of the FE from which it is supplied would not “lead to any fragmentation or dispersion of fiscal competence………Rather, such a solution would simply result in subjecting tour operators’ services (seen as a whole) to VAT at the place where they are actually provided to the consumer”. 

(4) He referred to a text book on EC tax law (P. Farmer and R. Lyal, EC Tax Law, Oxford, 1994, p. 160) where the authors noted that Berkholz “might be understood as expressing a reluctance on the part of the Court to refer to secondary establishments”.  In their view, however, the court’s words must “be read in the light of the circumstances of the case, in which a taxable person sought to escape the Community’s tax jurisdiction by creating national establishments outside Community territory”.  They submitted that, “in a genuine case in which a supplier ... has several business establishments all capable of performing services, the most appropriate method of determining the place of supply…..... would be to identify the establishment of the supplier whose resources were primarily used for supplying the service”.

433.       At [35] he concluded that the view put forward by DFDS to the contrary “is not in conformity with those principles - in fact it errs towards formalism”.  It failed to take account of the fact that “the economic realities of this case justify making travel agency business subject to VAT at the place where the services are provided.” Finally he noted that the legislature decided not to adopt a proposal to amend the definition of FE in article 9 to apply to any fixed installation of a taxable person, “even if no taxable transaction can be carried out there.” At [37] he said that the fact that the legislature chose not to do that was explained “precisely by the intention to emphasize the importance of the concept of the [FE].”

“That concept is an eminently economic concept.......It refers solely to an establishment from which services may be provided - by virtue of the sufficiency of the human and technical resources assigned to it - and are actually provided”.

434.        In conclusion, therefore, supplies such as those in issue were subject to VAT in the country where the FE was located:

“……provided that the company acting as agent is not autonomous and independent from the tour operator but is a mere auxiliary thereof and is in a form, which includes both human and technical resources, such that it is able to provide the services in question”.

435.       In the CJEU, at [17], the court similarly noted that article 26(2) uses the same concepts as those used in article 9 and it was, therefore, appropriate to refer to the rules arising from that definition of the place of supply.  The CJEU endorsed, at [18] to [20], the passages in Berghloz set out above.  They then proceeded to take a similar approach as the Advocate General had but with less detailed reasoning.

436.       They noted, at [21], that to treat all the services provided by a tour operator, including those supplied in other member states through undertakings operating on its behalf, as being supplied from its BE, would have the clear advantage of having a single place of taxation for all the business of that operator covered by article 26.  However, at [22], they said that that treatment would not lead to a rational result for tax purposes in that it took no account of the “actual place where the tours were marketed which, whatever the customer’s destination”, the national authorities have good reason to take into consideration as the most appropriate point of reference.

437.       They continued, at [23], that as the Advocate General pointed out at [32] to [34] of his opinion, “consideration of the actual economic situation is a fundamental criterion for the application of the common VAT system”.  The alternative approach for determining the place of taxation of the services of travel agents, based on the FE from which these services are supplied, is specifically intended to take account of the possible diversification of travel agents’ activities in different places within the Community:

“Systematic reliance on the place where the supplier has established his business could in fact lead to distortions of competition, in that it might encourage undertakings trading in one Member State to establish their businesses, in order to avoid taxation, in another Member State which has availed itself of the possibility of maintaining the VAT exemption for the services in question.”

438.       In those circumstances, at [24], it was concluded that, where services have been provided by a tour operator from a FE which that operator has in a member state other than that in which he has a BE, the supply of services to the customer is taxable where that FE is located.

439.       At [25], the court said that to determine whether, in these circumstances, the travel agent actually has such a FE in the relevant member state, it is necessary first to ascertain whether or not the relevant company “is independent” from the travel agent.  In that regard, at [26]:

“the fact that the premises of [the English company], which has its own legal personality, belong to it and not to DFDS is not sufficient in itself to establish that the subsidiary is in fact independent from DFDS.  On the contrary, information in the order for reference, in particular the fact that DFDS’s subsidiary is wholly owned by it and as to the various contractual obligations imposed on the subsidiary by its parent, shows that the company established in the United Kingdom merely acts as an auxiliary organ of its parent.”

440.       At [27], they continued that it is necessary to verify whether, in accordance with the case-law, the establishment in question is of the requisite minimum size in terms of necessary human and technical resources.  At [28] they concluded it was apparent from the facts set out in the order for reference, “particularly as regards the number of employees of the company established in the United Kingdom and the actual terms under which it provides services to customers, that that company does display the features of a [FE].”

441.       At [29] they concluded that the answer for the national courts therefore was that article 26(2) of the Sixth Directive is to be interpreted as meaning that:

“where a tour operator established in one Member State provides services to travellers through the intermediary of a company operating as an agent in another Member State, VAT is payable on those services in the latter State if that company, which acts as a mere auxiliary organ of the tour operator, has the human and technical resources characteristic of a [FE].”

Planzer Luzembourg and RAL

442.       HMRC also relied on the cases of Planzer Luxembourg and Case C-452/03 RAL (Channel Islands) Ltd and others v Commissioners of Customs & Excise.

443.       Planzer Luxembourg is not directly in point in that it relates to whether a Luxembourg haulage company, which had its headquarters in Switzerland, was entitled to a refund of VAT paid on fuel purchases in Germany.  The issue was whether the company had a place of business in Luxembourg (as required for it to make a reclaim) but the CJEU considered the meaning of FE in the course of considering that.  The company argued that it had such a place of business on the basis that it had in Luxembourg an office, two managers who carried on their activities in Luxembourg, five employees who worked part time as drivers from there, a telephone line and that goods vehicles registered in Luxembourg were used for the transport of air freight and invoices were drawn up at its offices there.  

444.       The court said, at [54] that it was well established that the term FE requires a “minimum degree of stability derived from the permanent presence of both the technical and human resources for the provision of the services (referring to Bergholz at [18], DFDS at [20] and ARO Lease at [15] and citing ARO Lease at [16] in full).  The CJEU said that concerning transport activities in particular, the term FE implies:

“at least an office in which contracts may be drawn up and daily management decisions taken and a place where the vehicles used for the said activities are stored……By contrast, registration of those vehicles in the member state concerned is not an indicator of a [FE] in that member state.”.

445.       In RAL a Channel Islands company, CI, operated slot machines in the UK in premises and using slot machines leased from a group company.  CI subcontracted the day to day management of the machines to another UK group company.  That UK company employed almost all of the group’s staff and CI had no staff of its own in the UK.  The functions which CI carried out directly were mainly confined to accounting and monitoring the cash flow from the machine.  The question was whether CI had a FE in the UK as a result of the presence of the slot machines

446.       At [39] and [40] the Advocate General interpreted Berkholz, as being a case decided on the basis that there was simply no staff assigned on a permanent basis on the ferry.  He said, at [41]:

“The court only required the presence of a “minimum size” of establishment and no more and no less than the resources “necessary” for the provision of the services of a permanent nature.  The Court did not make the permanent presence of all possible human and technical resources, possessed by the supplier himself, in a certain place, a precondition for adoption of a minimum-requirements test for characterising a given set of circumstances as constituting a [FE] …….which was subsequently followed and adopted by the Court, in particular in ARO Lease and DFDS.”

447.       Having summarised the decisions in ARO Lease and DFDS he noted, at [44], that the court expressly affirmed in DFDS that “consideration of the actual economic situation is a fundamental criterion for the application of the common VAT system”.  His view was that “here it is necessary to undertake an analysis that is especially responsive to the factual economic and commercial reality of the case.”

448.       He continued, at [45] and [46], that in view of these cases, in the present case, the supply of gaming services from the slot machines was made from FEs in the UK.  He noted that the arcade where the machines were located had regular opening hours, like any other business establishment, and there were staff permanently attending to customers and looking after the premises and machines.  He considered this human element was important in distinguishing the case from the situation in Berkholz.  Such “a permanent human presence on the premises lends stability to CI’s supply of slot gaming services in the amusement arcades” and “most importantly, these [FEs] are not on board sea-going vessels moving from one country to another”, a circumstance which could justify the option in favour of the BE.

449.       He said, at [47], that the problem was whether the FEs should be considered as CI’s FEs.  He noted, at [48], that CI argued that its only presence in the UK took the form of the leased slot gaming machines and essentially “in order for it to have a [FE] in the UK it would have to possess all the necessary human and technical resources itself there”.  He disagreed noting that ARO Lease and DFDS cases were particularly enlightening. 

(1) At [49] he took from ARO Lease that it was not “absolutely indispensable that the persons working in the amusement arcades be CI’s own staff” and the necessary “structure” will “inevitably vary depending on the sector concerned”.  He took the view that the court in ARO Lease indicated that the presence of staff or of a framework for making contracts or premises were alternatives; not all of these were required for there to be a FE (see [42]).

(2) He continued, at [50], that in DFDS, the English company had its own legal personality distinct from that of its Danish parent but the court nevertheless said it was tantamount to a FE of the parent.  The parent itself did not have employees in the UK and did not own premises there. However, the parent had obtained through contractual arrangements with the English subsidiary, acting as its agent, the human and technical resources to supply its tour services in the UK and the company was merely an auxiliary organ of its parent. 

450.       He considered, at [52], there was an essential distinction to be made between those resources which “necessarily have to be under the direct dependence of the supplier in a certain place for a FE to be his” and resources which do not have to be under the direct dependence for a FE to belong to the supplier albeit they may confer a fixed character on an establishment.  He said that those which must be under the direct dependence of the supplier are “resources directly involved in the supply of the particular service in question, namely the conclusion and performance of the contracts with customers, necessary for the supply”.

451.       He said, at [53], that in reality, to demand, as CI argued, that the persons whose presence is an important factor in conferring a fixed character on an establishment “must all be employees or directly dependent upon the supplier would lead to absurd results”.  He gave an example were there are security staff who are the only people with keys to the establishment and are in charge of opening and closing the premises at regular hours. Such persons are certainly indispensable to ensure that the establishment does not operate merely intermittently. They should be considered “as human resources whose permanent presence is necessary for the provision of the services in the establishment to take place and, therefore, to confer a fixed character on the establishment”.  He said “it would certainly be unacceptable that such an establishment would cease to be characterised as a FE of the supplier of the services by virtue of the fact that he had decided to outsource the activities of security in the establishment to an independent security company”.

452.       He continued, at [54], that the relevant staff in this case did mainly practical tasks (such as the provision of music, refreshments and change to customers, emptying the machines’ cash boxes, witnessing large payouts, providing security, carrying out maintenance and so forth).  Such activities were “ancillary” with respect to the supply of the slot gaming machine services. 

453.       He noted, at [55], that the staff did not have any direct involvement in the conclusion of the gaming contracts between CI and the customers. Rather the supplies made in this specific sector are based on “discrete contractual arrangements made between each customer and CI directly through the slot machines themselves”, which were “concluded and performed entirely in the UK every time a customer inserts a coin in a slot gaming machine operated by CI”.  On that analysis, it was the machines themselves, as automatic devices, which enabled CI to supply the slot gaming services directly to each customer in the UK.  In this specific sector, therefore, “the slot gaming machines are the crucial and sole structure in the ‘amusement arcades’ that has to be under the direct dependence of CI to allow the conclusion that each of those ‘amusement arcades’ where its slot machines are installed is a FE”.

454.       He said, at [56], that the activities actually performed by CI in Guernsey did not seem to be a decisive feature of the supply of the relevant gaming machine services to each customer, which occurred automatically as he had described.  He concluded, at [57], that a company, such as CI, which supplies gaming machine services directly to its customers through leased gaming machines it operates in premises in the UK, with the aid of auxiliary staff outsourced from third companies to perform ancillary activities necessary to confer a permanent character on the supply, should be regarded as having a commercial structure in the UK with the minimum resources required for it to be considered a FE.

455.       He then considered, at [58], whether, reference to the BE rather than the FE would be “rational for tax purposes”.  He referred to the comments in DFDS on the potential for distortion of competition.  He considered, at [62], that in Berkholz and Faaborg-Gelting:

“the Court took clearly into consideration the decisive aspect that the subjection to the VAT system of the slot gaming machine and restaurant services was not at risk in the particular circumstances of the two cases.  If the place where the suppliers had decided to establish their place of business had in those cases been located outside the territory of the Community, the application of that connecting factor would certainly have raised many doubts.”

456.       He continued, at [63], to note that the result of taxing by reference to the BE in this case would be that the slot-gaming machine services provided in the UK to UK consumers would not be taxed at all (as CI was based in Guernsey).  He concluded, at [64]:

“In the present case, in contrast to the position in DFDS, there is not merely a risk of prompting companies to establish their places of business in Member States that are able to maintain more favourable VAT regimes for the services in question.  The risk in the present case is that of encouraging companies to relocate and establish their businesses outside the VAT territory of the Community, while continuing to supply their services in that territory in [FEs]….to consumers residing there.”

Caselaw on current place of supply rules - Welmory

457.       These principles have been considered more recently in the context of the revised place of supply rules, applicable under articles 44 and 45, in C-60/125 Welmory sp. z o.o. v Dyrektor Izby Skarbowej w Gdańsku [2015] STC 515.  The facts, as set out in the decision by the CJEU, were as follows.

(1) In 2009, a Cypriot company (“C”) entered into a cooperation agreement with a Polish company (“P”) whereby (a) C agreed to provide P with the service of making available an internet auction site and associated services relating to the leasing of the servers needed for the site to function and the display of the goods to be auctioned; and (b) P undertook principally to sell goods on that site.

(2) The customer first purchased a number of bids from C on the online site.  The bids purchased gave the customer the right to take part in the sale of the goods offered for auction by P on that site and to make an offer to purchase one of the items.  The goods were sold to the customer who by means of his bids offered the highest price for them.

(3) The source of P’s income was (a) the selling price for the goods obtained in the online auctions and (b) remuneration received from C corresponding to part of the proceeds of sale of the bids.

(4) On 19 April 2010 C acquired 100% of the share capital of P.

(5) For the period from January to April 2010, before that acquisition, P issued four invoices for services supplied to C such as advertising, servicing, provision of information and data processing.  The issue was the correct VAT treatment of these services.

458.       P took the view that its services were supplied at the BE of C in Cyprus.  However, the Polish tax authorities considered that these were supplies of services to a FE of C in Poland which should be taxed in Poland.  The court in Poland upheld that decision essentially on the basis that the two companies’ activities formed an economically indivisible whole, as the object of their entire business could be achieved in Poland only through cooperation between them.  They noted that C made use in Polish territory of P’s technical and human resources, so that P was to be treated as C’s FE in Poland.  On a further appeal, the matter was referred to the CJEU.   

459.       The Advocate General noted, at [23], that VAT “is in principle owed to the [state] in which the service is consumed” and “as a rule this is likely to occur at the place where the recipient is established”.  Therefore article 44 allocates the power to tax to the state of the recipient of the service.  She noted, at [26], that, as the intention was that the place of supply rules should be applied uniformly to determine unequivocally the right to tax, the interpretation of the provisions “must in any case guarantee legal certainty in order to avoid conflicts” between member states on jurisdiction.

460.       She concluded from this that the determination of a FE within the meaning of article 44 “thus requires first and foremost that criteria are practicable” and, at [27], that “an unreasonable administrative burden on taxable persons should also be avoided through a flexible approach to determining who is liable for tax”.  In that context, at [29], she said that for the service provider there must be legal certainty as to the existence of a FE of the recipient of his service and, at [30], service providers cannot be expected either to carry out extensive investigations into the recipient or to put up with uncertainty on the issue of their liability for tax.  For this reason too there “is a need for objective and clear criteria”.

461.       The Advocate General then referred, at [31], to the case law on the meaning of FE noting that to date that had mainly taken place in regard to article 9(1).  She described it, at [32], as settled case-law that, as regards article 9, the place of the BE is the primary point of reference and, at [33], as established that a FE:

“only exists if the taxable person’s place of business has a sufficient degree of permanence and a structure adequate, in terms of human and technical resources, to supply the services in question on an independent basis”. 

462.       The Advocate General referred, at [34], to ARO Lease and the fact that the court held that it was significant in the vehicle leasing business whether a taxable person had its own staff available to it at a given place and whether contracts could be drawn up there or decisions made on the management of the business.

463.       She referred to DFDS, at [35] as deciding that “a company which, although having its own legal personality, is completely controlled by its parent company may be regarded as a [FE] of the parent company”.  She continued, at [36], that the judgment in DFDS was not of general application as it only interpreted the special tour operators’ rules:

“Although in its reasoning it also referred to the general rule for determining the place of supply of a service, the judgment in DFDS is not, however, capable of general application, as the Court recently found in the judgment in Daimler.  Furthermore it serves the purpose of legal certainty in regard to the person liable for tax if a legal person with its own legal personality cannot at the same time be the [FE] of a different legal person.”

464.       She noted, at [37], that these case-law principles in relation to article 9(1) related only to the question of when the provider of a service maintains a FE and not the determination of a FE of the recipient of a service.  She thought, at [38] to [42], however, that the case law remained relevant for the reasons she set out (which were essentially endorsed by the CJEU (see below)). 

465.       At [43] as regards the current definition she queried whether a FE within the meaning of article 44 is required not only to use services but also to be capable of performing its own taxable supplies.  However, that did not need to be decided because, provided C maintained a FE in Poland, this establishment would also carry out services in the form of the operation of the auction website.  Moreover:

“from a factual point of view it is doubtful whether as a rule every structure which, in terms of its human and technical resources, is able to use services for its own needs would not indeed at least have the possibility of supplying services itself…”.

466.       At [44] she said that ultimately, for reasons of legal certainty, the precedence in consistent case-law given by the court to the BE should also apply to article 44.  It serves the requirement for legal certainty if the place of supply of the service is, in case of doubt, linked to the BE of the recipient of the service, which as a rule is a more easily ascertainable objective criterion than the maintenance of a FE.  She also considered this was consistent with the relationship, expressed in the wording of article 44, between the basic rule in the first sentence relying on the BE and the exception in the second sentence concerning a FE.

467.       She then gave her view on the application of these principles to the facts of the case.  She noted, at [45], that a FE of C in Poland:

“can thus be assumed only if that company has an establishment there which displays a sufficient degree of permanence and a suitable structure in terms of human and technical resources to enable it to receive and use the services supplied to it for its own needs.  In case of doubt, the assumption is that no [FE] exists so that pursuant to the first sentence of [article 44] the Cypriot company’s place of business is to be regarded as the place where the service is supplied.”

468.       She noted, at [46] and [47], that ultimately it was for the referring court to make a finding on the basis of the facts but in making its findings, the referring court must take account of the points she then set out.  First, at [48], for there to be an FE in a member state:

“it is not necessary for the taxable person to have at his disposal there human resources who are employed by him, or to have technical resources which he owns.”

469.       In that regard, she noted, at [49], that it had previously been held that it would lead to unacceptable results if a FE were to be assumed only where the human resources deployed are employed by the taxable person himself.  It would also invite abuse if a taxable person were able to transfer the taxation of his services from one state to another simply by covering his human resource requirements using different service providers.  The court’s reference, when determining a FE, to the absence of human resources in a vehicle leasing business, in her view, were not to be applied generally.

470.       Accordingly, she considered, at [50] to [51], that a FE of C in Poland was not ruled out “merely because it used [P’s] technical resources and human resources not employed by it”.   However, she explained that did not mean that P would function simultaneously as both the service provider and, as the FE of C, as the recipient of the service.  This is because:

“even if a [FE] does not necessarily require its own human and technical resources, the taxable person must nevertheless - based on the requirement for a sufficient degree of permanence in relation to the establishment - have comparable control over the human and technical resources.  Therefore employment and lease contracts are required in particular in relation to the human and technical resources which put the latter at the taxable person’s disposal as if they were his own and which therefore also cannot be terminated at short notice.”

471.       She continued on this point, at [52] and [53], that in other words, it should be emphasised that a taxable person cannot as such constitute a FE of a different taxable person.  This does not, however:

“exclude the possibility of a taxable person having immediate and constant access to the human and technical resources of a different taxable person who, in a different respect, can at the same time be a service provider for the [FE] thereby constituted.

That said, where the human and technical resources of the service provider and of the establishment of the recipient of a service are virtually the same, it may be questioned whether there is a supply of a service to another taxable person at all.”

472.       She noted, at [54], that the UK had “rightly” referred to the fact that the question where C conducted a business activity and used P’s services was also crucial.  In order to be “fixed” within the meaning of the article 44, the establishment must be capable of using services for its own needs.  She said, at [55], that the referring court would have to examine what independent business C carried out using the human and technical resources at its disposal in Poland and whether P’s services under the cooperation agreement were applied specifically for this business.

473.       She concluded, at [56], that the answer to the question referred was, therefore, that a FE within the meaning of article 44 is:

“an establishment characterised by a sufficient degree of permanence and a suitable structure in terms of human and technical resources to enable it to receive and use the services supplied to it for its own needs. It is not necessary for it to have its own human and technical resources for this, provided the third-party resources at the establishment are available to it in a way that is comparable to having its own resources.”

474.       As noted, the CJEU took the view that the court’s case law on article 9 continued to be relevant.  They noted, at [43] to [47], that the wording of article 44 is similar to that of article 9, that those provisions both determine the point of reference for tax purposes of supplies of services and pursue the same objective, so that the case law on the interpretation of article 9 “can in principle be applied mutatis mutandis to the interpretation of [article 44]”.  They considered that conclusion was borne out by the Regulation, noting it was apparent from recital 14 in the preamble “that the EU legislature wished to clarify certain concepts necessary for determining criteria relating to the place of taxable transactions, while taking account of the relevant case-law of the court”. Even though the Regulation was not yet in force at the material time, it should none the less be taken into account. 

475.       Turning, therefore, to the cases relating to article 9, they noted, at [53] and [54], that it is settled case-law in that context that the primary point of reference is the place where the taxable person has established his business and they said that should also apply as regards article 44 (referring to Bergholz, Faaborg-Gelting and ARO  Lease).  At [55], they said that, as was the case under article 9, that provided greater legal certainty:

“the place where the taxable person has established his business as primary point of reference appears to be a criterion that is objective, simple and practical and offers great legal certainty, being easier to verify than, for example, the existence of a [FE].  Moreover, the presumption that the services are supplied at the place where the taxable person receiving them has established his business makes it possible both for the competent authorities of the Member States and for suppliers of services to avoid having to undertake complex investigations in order to determine the point of reference for tax purposes.”

476.       They noted that furthermore, at [56], the place of business is mentioned in the first sentence of article 44, whereas the FE is mentioned only in the following sentence. That sentence, introduced by the adverb “however”, “can only be understood as creating an exception to the general rule set out in the previous sentence”. 

477.       At [58], the court referred to formulation of the FE test set out in Planzer Luxembourg, C-73/06, at [54] and the case-law cited there, which “directly inspired” the wording of article 11 of the Regulation (see [444] above).  So, at [59], for C to have a FE within the meaning of article 44 it:

“must have in Poland at the very least a structure characterised by a sufficient degree of permanence, suitable in terms of human and technical resources to enable it to receive in Poland the services supplied to it by the Polish company and to use them for its business, namely running the electronic auction system in question and issuing and selling ‘bids’.”

478.       They said, at [60], that the fact that the business such as that carried on by C, consisting in operating a system of electronic auctions which comprised making an auction website available to P and issuing and selling bids to customers in Poland, could be carried on without requiring an effective human and material structure in Polish territory was not determinative.  Despite its particular character, such a business requires at least a structure that is appropriate in terms especially of human and technical resources, such as appropriate computer equipment, servers and software.

479.       At [61] they noted that P argued that in fact the infrastructure it made available to C did not enable C to receive and use for its business the services supplied to it by P.  It argued that the human and technical resources for the business carried on by C, such as computer servers, software, servicing and the system for concluding contracts with consumers and receiving income from them, were situated outside Polish territory. 

480.       The court said, at [62], that it was for the national court to verify such factors but, at [63], if the facts alleged by P were shown to be correct, the referring court would then be led to conclude that C did not have a FE in Poland, since it did not have the necessary infrastructure to enable it to receive services supplied by the Polish company and to use them for its business.

481.       The court continued, at [64], that:

“the fact that the economic activities of the two companies, which are linked by a cooperation agreement, form an economic whole and that their results are of benefit essentially to consumers in Poland is not material for determining whether [C] possesses a [FE] in Poland.. …the services supplied by [P] to [C] must be distinguished from those supplied by [C] to consumers in Poland.  They are distinct supplies of services which are subject to different schemes of VAT.”

482.       At [65] the court concluded that a person has a FE within the meaning of article 44, for the purpose of determining the place of taxation of the relevant services, if that establishment is characterised by a sufficient degree of permanence and a suitable structure in terms of human and technical resources to enable it to receive the services supplied to it and use them for its business, which is for the referring court to ascertain.

Discussion on FE issue – overview of the parties’ positions

Overview of FE test

483.       It is established in the above cases, as now reflected in the Regulation, that there are two inter-related elements to what is required for there to be a FE, through which a business makes supplies or, at which it receives supplies:  

(1) The establishment must be of a minimum size with both the human and technical resources “necessary”,  “adequate” or “suitable”:  

(a)    for the provision of the supplies on an “independent basis”, when looking at the location of the supplier, or

(b)   to enable the establishment to receive and use the relevant supplies for its own needs, when looking at the location of the recipient of the supplies.

(2)  As stated in Bergholz, those resources must be permanently present or, as phrased in ARO Lease, the establishment must have “a minimum degree of stability derived from the permanent presence of both the human and technical resources” necessary for the provision of the supplies or to enable the establishment to receive and use the relevant supplies for its own needs. 

484.       The parties disagreed as to how each part of the test is to be interpreted in the context of both determining (a) the place where supplies of insurance were made by Advantage and (b) where the services were supplied by the appellant to Advantage. 

Supplies made by Advantage – parties’ position on permanence or control requirement

485.       In summary, the appellant argued that the permanence aspect of the FE test was satisfied only if Advantage controlled the relevant resources of the appellant as if they were its own by reference to the comments of the Advocate General in Welmory (see, in particular, [468] to [471] above).  In its view that was plainly not the case; essentially the appellant and Advantage operated as two separate commercial enterprises with different functions.

486.       In HMRC’s view, the Advocate General’s “control” test in Welmory is out of line with DFDS and the other authorities.  They argued that it suffices that the appellant’s relevant resources were available to Advantage under the contractual arrangements on an on-going stable basis and that the appellant was required to conduct its business for Advantage in ways which were approved by Advantage.  HMRC considered that common ownership was not necessary for one company’s resources to be a FE of another.  However, common ownership may reinforce any contractual links to establish that there is a FE.  The long term service agreements and close relationship between the parties, combined with the common ownership links, were enough to satisfy the test.  As regards contractual links, HMRC noted that Advantage retained control of the risks which could be insured, the policy wording, the net rates and ultimately over claims handling (albeit that the appellant dealt with the “customer facing” side of things). 

487.       The appellant did not dispute that there is a temporal aspect to the permanence test or that the on-going contractual relationship between the parties sufficed to meet that aspect of the test.  However, in its view HMRC largely ignored the qualitative “control” part of the test.  The appellant considered that for the level of control required to be comparable to that of an owner is entirely consistent with the previous case law.  In RAL, for example, the Advocate General referred to the need for the relevant resources to be under the “direct dependence” of the business in question.  Whilst in the appellant’s view the courts in DFDS approached the issue wrongly by looking at whether the English subsidiary was independent of the parent according to the approach in agency cases, they were clearly concerned with a high level of control in making that assessment. In any event, the circumstances of this case are very different from those in DFDS.

Supplies made by Advantage – parties’ position on resources necessary

488.       The appellant argued that, in any event, the appellant’s resources in the UK did not provide Advantage with all that was necessary for it to make insurance supplies to UK customers from a UK establishment on an independent basis.  It is clear, from Welmory in particular, that this test is not satisfied if the relevant resources do not provide key functions required to make the relevant supplies.  In fact, Advantage used its own resources in Gibraltar for a number of the key functions required to make supplies of insurance such as the underwriting decisions, dealing with large losses and dealing with the reinsurers, the investment of funds and the regulatory aspects.

489.       HMRC responded that, relying in particular on ARO Lease, DFDS, RAL and Planzer Luxembourg, it suffices that the appellant’s human and technical resources provided a framework comprising all that was necessary for the “customer facing” side of the business to be dealt with in the UK without reference to Advantage’s BE in Gibraltar.  The appellant clearly had the necessary human and technical resources necessary for the provision of these services given the large scale of its staff and the fact that it owned or controlled the relevant technical resources (see [142] above).  These resources provided a framework which enabled the services to be supplied on an independent basis given that the appellant’s staff provided quotes for insurance, concluded contracts on behalf of Advantage within the underwriting guidelines set by Advantage, issued the insurance policies and accepted payment, handled all claims as regards contact with customers and made daily management decisions as regards customers.   

Discussion - supplies of insurance made by Advantage

490.       Whilst the tests set out in the cases are an interrelated whole, we have treated the different aspects separately in line with how the parties approached this.

Economic considerations

491.       There was some debate as to the relevance of and stage in the analysis at which economic considerations should be taken into account.  The appellant submitted that such considerations are not to be taken into account in determining whether there is a FE in the first place (referring to ARO Lease at [23] and Welmory at [64]).  In its view, they are relevant only once it is established that there is a FE in a different place to the BE in then applying the rationality “tie breaker” test to decide between the two places as the place of supply/belonging.  HMRC noted that the courts have generally addressed, as the first issue, whether it is appropriate to look away from the location of the BE on the basis it does not give a rational result. 

492.       In our view, it is immaterial whether we first determine whether there is a FE or whether it is in any event appropriate to look away from the location of the BE as the place of supply/belonging.  The issue only arises, however, if there is in fact a FE such that it seems logical to address that first.  We have considered the extent to which economic considerations are relevant to that question in the discussion set out below. 

Permanence or control requirement

493.       There is little elaboration in the earlier cases as to precisely what is required for there to be a FE and when it is permissible to look beyond the BE.  In Bergholz it appeared to be the lack of human resources required to maintain the gaming machines on board the ferry which meant there was no FE.  The limited intermittent presence of staff (for two hours a day) did not suffice particularly, the court said, where the BE provided an appropriate point of reference (see [407] and [408] above). 

494.       The appellant submitted that whilst Faarborg-Gelting contains very little in the way of reasoning, the court must have concluded that the restaurant on board the ferry was not a FE, due to the lack of the resources on the ferry required for key management functions (such the resources to make decisions on the menus, pricing and for negotiating and contracting with suppliers).  Clearly the resources required to run the restaurant on a daily basis were present on the ferry.  We note, however, that there is no real indication the court considered such factors or indeed to what extent they considered whether there was a FE; they referred rather to the fact that the BE provided an appropriate point of reference (see [410] above). 

495.       As HMRC submitted, in both Faarborg Gelting and Bergholz, the court seemed to be influenced by the fact that treating the gaming machine or restaurant located on a ferry as a FE would have created difficulties because the ship was going through different territorial and non-territorial waters (as we note the Advocate General in RAL also suggested).   

496.       We find more guidance in ARO Lease (see [411] to [418] above) where the CJEU sought to explain the Bergholz formulation as meaning that an establishment could be regarded as a FE from which services are supplied only if:

(1)  it has a “minimum degree of stability derived from the permanent presence of both the human and technical resources necessary for the provision of the services”;

(2) and, as it was later put, it has “a sufficient degree of permanence and a structure adequate, in terms of human and technical resources, to supply the services in question on an independent basis.” 

497.       Applying that test, the CJEU concluded that the Netherlands car leasing company did not have a FE in Belgium as the result of the activities of self-employed intermediaries there as it did not have there either its own staff or a structure which “has a sufficient degree of permanence to provide a framework in which agreements may be drawn up or management decisions taken and thus to enable the services in question to be supplied on an independent basis”.  That was on the basis that the services supplied consisted “principally in negotiating, drawing up, signing and administering the relevant agreements” and in making the vehicles physically available to customers.

498.       We understand from the CJEU’s formulation of the test, as applied to the facts of that case, that human and technical resources in a particular location will potentially have the minimum degree of stability or fixed quality to comprise an entity’s FE only if:

(1) The resources provide what is necessary or adequate for the entity to make the relevant supplies at that location on a sufficiently stand-alone basis, meaning without reference or, at any rate, substantial recourse, to the resources of the BE.

(2) As suggested by the reference to permanence and stability, those resources are available in a sufficiently certain way that they can potentially be utilised by the entity at that location for that purpose on an on-going basis. 

499.       Clearly part of the “permanence” test is a temporal requirement.  Whilst we do not interpret “permanence” to require an indefinite continuation, broadly speaking, it seems the resources must not be merely temporarily available or capable of being withdrawn at short notice.   As noted, there was no dispute that any such temporal requirement was satisfied in this case, as a result of the on-going nature of the contractual relationship between the parties.

500.       The references to “permanence” and “stability” also carry the implication that for there to be sufficient certainty as to the on-going or non-temporary nature of the establishment the resources need to be under the ownership or, at any rate, under some degree of control of the relevant entity.  Hence the conclusion in that case that the self-employed intermediaries “who bring interested customers into contact with [ARO Lease]” could not be said to be human resources of permanence within the meaning of the case-law. 

501.       The terms of reference used, such as a “minimum degree of stability” and “a sufficient degree of permanence and a structure adequate” indicate there is no bright line test as to what suffices for an establishment to be a FE. It is a question of what suffices to provide a minimum, sufficiency or adequacy in the context of the business in question and the nature of the particular supplies. 

502.       We note that the principle that the relevant assets must be owned by the business or at least under some degree of its control follows in any event simply from the fact that the FE has to be a FE of the relevant entity from which it makes supplies (or receives supplies).  The FE test essentially comes into play as regards establishing if a taxable person may be providing (or receiving) supplies at a different establishment from its BE.  It is the place of the operation of its business, in the context of its ability to make or receive supplies, which is in question. 

503.       The Advocate General in RAL made this point when he spoke of the need for the resources necessary for the provision of the relevant supplies, in the sense of those directly involved in the making of the supplies, to be under the “direct dependence” of the relevant entity in order for a FE to be its FE (see [445] to [454] above).  In that case, he considered that the resources in the UK which were directly involved in the supplies of gaming services made by a Guernsey company comprised only the leased gaming machines themselves, as the medium through which contracts for the provision of the gaming services were concluded and performed.  There were staff in the UK who he considered contributed to the fixed character of the establishment; that was the essential distinction with Bergholz.  However, they carried out ancillary functions which were not directly involved in the supplies of gaming services.  What mattered, therefore, was that the gaming machines were under the “direct dependence” of the Guernsey company (as they were, as its leased assets) and not whether the staff were.

Welmory and DFDS approaches

504.       The ownership or control issue was thrown into focus in DFDS and Welmory where, as here, it was argued that an entity had a FE in a different country, in effect, through access to the resources of a separate legal entity.  It was accepted in both cases that an entity does not actually need itself to own or employ the relevant resources for it to have a FE as a result of those resources.  However, a different approach was taken to assessing what level of control over the entity/its resources sufficed and what links between the two entities were relevant to that assessment. 

505.       In DFDS the Advocate General and the CJEU approached this in terms of assessing whether the English subsidiary, which owned the relevant resources, acted independently from its Danish parent (see [419] to [441] above).  The CJEU considered that the fact that the premises of the English company, which had its own legal personality, belonged to it was not sufficient in itself to establish that the subsidiary was independent from DFDS.  The fact that the subsidiary was wholly owned by the parent and the various contractual obligations imposed on it by its parent, showed that it merely acted as “an auxiliary organ of its parent.”  The Advocate General, relying on cases on whether an agent acts independently, noted that ownership of the subsidiary indicated its dependence on the parent but pointed also specifically to the fact that the subsidiary took no financial risk in relation to the contracts with customers, it could only carry on business for others with the consent of the parent and that it required approval from the parent as regards a number of management matters. 

506.       In DFDS the parent company was held to make supplies of package tours through its English subsidiary, as its FE, therefore, on the basis that the subsidiary was merely, as suggested by the term “auxiliary”, assisting the parent in making the relevant supplies subject, in all important respects, to the parent’s control, without the ability to conduct other business without the parent’s say so or to take its own important management decisions and without taking its own financial risk.  As was indicated by the fact that it was a subsidiary of the parent and confirmed by the other factors, it was entirely subordinate to the parent.  The courts concluded that the subsidiary otherwise qualified as a FE, broadly, due to the large numbers of staff employed by it in the UK and as it initiated the contracts with customers in the UK. 

507.       In Welmory, the Advocate General thought that it was not in accordance with principles of legal certainty for a taxable person itself to be a FE of a different taxable person.  She accepted, however, that did not preclude the possibility that its resources may themselves comprise a FE of the other person.  She interpreted the “permanence” requirement as meaning that would be the case, in that context, only if the other person had “control” of the resources of the first person as if they were its own resources, in the sense of having “immediate and constant access” to them.  She said that, therefore, “employment and lease contracts are required in particular in relation to the human and technical resources which put the latter at the taxable person’s disposal as if they were his own and which, therefore, also cannot be terminated at short notice.” (See [467] to [471] above.)

508.        The CJEU in Welmory did not refer specifically to the Advocate General’s comments on this.  They simply said that a FE must be characterised by a “sufficient degree of permanence” referring to the general Bergholz formulation of the FE test as set out in Planzer (which cross referred to a similar general statement in DFDS (at [20]) (see [477] above).  They focussed rather on whether the structure was suitable to enable C to receive the relevant services in Poland and to use them for its business (see [470] to [480] above).  They then rejected the rather broad brush position put forward by the Polish authorities (see [458]).  They held that the fact that the economic activities of the two companies, which were linked by a cooperation agreement, formed an economic whole and that their results were of benefit essentially to consumers in Poland was not material for determining whether C had a FE in Poland (see [481] above).  They noted that each set of supplies (of services by P to C and by C to consumers) had to be assessed separately. 

509.       The appellant argued that the CJEU can be taken to have endorsed the Advocate General’s view on “control” in these comments.  Its view was that, if such a close economic link does not suffice, the circumstances where there is the required level of control for there to be an FE must be exceptional (as, in the appellant’s view, accords with the general precedence to be given to the BE).  HMRC thought the CJEU was merely making a point that the two sets of supplies had to be viewed separately, in the context of their comments on the adequacy of the structure.  They noted that the CJEU followed the approach in the other cases of looking at “permanence” and, in effect, decided the case on the point as to the adequacy of the resources available to C in Poland.  As noted, they considered that the Advocate General’s approach is out of kilter with the general approach in the cases and, in particular, that in DFDS.

510.       We note that the CJEU did not focus on the permanence requirement and did not specifically endorse the Advocate General’s comments on control, but they did not disapprove of them either.  We do not think the CJEU’s comments on the economic position can be taken to be an endorsement of the Advocate General’s views on control as such.  The two concepts are not necessarily one and the same.  For example, a business may work in co-operation with another with a view to a common business goal without one business having any control over the resources of the other, in the sense set out by the Advocate General.  However, the CJEU were clear that such an overall economic link did not of itself determine there was a FE (whether in respect of the permanence requirement or otherwise).  We have commented on this further in our conclusions.

511.       Whilst the Advocate General spoke in terms of control rather than the more usual terms of permanence or stability, as set out above at [500] to [503], in our view it is clear that a requirement for some element of control of the relevant resources is an essential element of the FE test.  Moreover, we do not consider that in framing the test in this way, as requiring control comparable to that of an owner, the Advocate General can be said, as HRMC argued, to be setting an unduly high hurdle compared with the other authorities.  In RAL and ARO Lease it was envisaged that the relevant resources had to be under the “direct dependence” of or actually to be employed by the relevant entity. 

512.       As regards DFDS, the appellant noted that the Advocate General in Welmory said that DFDS was not to be regarded as of general application given it relates to the TOMS rules and she relied on the Daimler case in support of that.  In the appellant’s view this reflects that the reasoning in DFDS, in looking at the concept of independent agency, is out of kilter with the general approach in the cases.  The appellant submitted that the FE test is not synonymous with an independent agency or permanent establishment test.  On that basis, the factors taken into account in DFDS by reference to those tests, such as whether a company is wholly owned and acted only on its parent’s behalf, are simply not relevant to the FE test (and again that approach is out of line with the approach in Welmory of disregarding the close economic link). 

513.       In support of this, the appellant referred to an opinion adopted on 19 September 2016 by the VAT Expert Group, a group which comprises academics and experts from member states and which was set up by the European Commission to assist and advise the Commission.  In this opinion, the group said that, as the Advocate General noted “correctly” in Welmory, “DFDS must be viewed as an exceptional case on its facts not…capable of general application”. They expressed concern that there is “increasing evidence that tax authorities incorrectly conclude that an FE exists on the basis of the existence of PE, and vice versa” and set out a number of problems they saw with that.

514.       The appellant submitted that the approach in DFDS is also out of kilter with the approach taken in the UK to outsourcing cases (such as Capital One Bank (Europe) Plc v Revenue and Customs Commissioners [2006] BVC 2148, [2005] STI 2042 and MBNA Europe Bank Ltd v Revenue and Customs Commissioners [2006] EWHC 2326 (Ch), [2006] STC 2089 at [121]).  In Capital One, for example, it was held that a Jersey company belonged in Jersey despite what was described as “considerable” outsourcing to a UK company and although it had “little substance” in Jersey (see [157] and [158]).

515.       The appellant concluded that, in any event, it is likely that DFDS would have been decided in the same way even on what it considered to be the correct Welmory formulation of the test due, in particular, to the finding that the subsidiary bore no financial risk in acting for the parent and was described as a mere auxiliary of the parent.  On that basis the parent may well have had control of the subsidiary’s resources akin to that of an owner but, on the evidence, the situation in this case is far from that in DFDS. 

516.       HMRC said that the approach in DFDS is the right one to follow as that is the only case in the CJEU in which the “permanence” requirement has been considered in detail.  The Danish parent only had access to certain key aspects of the English company’s operations but that sufficed for the subsidiary/its resources to be regarded as a FE of the parent having regard to ownership and the contractual constraints on the subsidiary.  They considered that there was no reason to regard DFDS as somehow of limited application.  They noted that an Advocate General’s opinion is only persuasive and not binding authority.  The CJEU in Welmory cross-referred to DFDS (as it was referred to in the passage of Planzer they referred to) such that it cannot simply be disregarded.  Moreover they submitted that the Daimler case, which the Advocate General referred to in support of her views, was actually on a different point and not really relevant. 

517.       As the appellant submitted, the approach adopted in DFDS is somewhat out of kilter with the other authorities in assessing whether the English subsidiary/its resources was a FE of the parent by reference to whether it was independent of the Danish parent.  However, whilst the Advocate General in Welmory cast doubt on DFDS as an authority of general application, there is no indication in the CJEU decision in that case, or the other CJEU decisions to which we were referred, that it is to be disregarded.  The Daimler case to which the Advocate General referred is, as HMRC submitted, on a different point. 

518.       In brief, in Daimler the issue was whether a company based in Germany could recover input tax incurred through a Swedish subsidiary which carried out winter testing of cars in Sweden but carried out no taxable activity. The input tax was recoverable under the relevant provisions only if Daimler did not have in Sweden “a [FE] from which business transactions were effected”. The court decided that it did not have such a FE primarily because Daimler was not actually making any supplies in Sweden.

519.        The CJEU rejected the Swedish government’s argument that, as noted at [47], there was a FE on the authority of DFDS, on the basis that the Swedish company was a subsidiary which had technical and human resources and acted as a mere auxiliary of the parent company.  They noted, at [48], that the subsidiary was a taxable person on its own account and that it had not actually incurred the input tax.  They described DFDS, at [49], as a case where the “independence of the status of the subsidiary was disregarded in favour of the commercial reality only to ascertain which of the parent company and the subsidiary had actually carried out the active taxable transactions of supplies of services” and so which member state had taxing rights.  They concluded, at [50], that in this case there was no FE because the testing department in Sweden did not actually carry out any active output taxable transactions. 

520.       Moreover, although the DFDS case relates to the place of supply rules in the context of the TOMS rules, those rules are framed in the same manner as the general rule we are concerned with.  We see no reason to regard the DFDS case as of limited application because it was decided in that particular context, albeit that, of course, its relevance in other cases has to be assessed by reference to the particular facts.  Overall, we consider we must have regard to the FE test in DFDS as the current case law in this area stands. 

Differences in Welmory and DFDS approaches

521.       In our view, the two approaches of assessing the independence of the entity/its resources, or the level of control over the underlying resources and functions of the entity, are not wholly irreconcilable. On its normal meaning, the term independent means free from outside control or not subject to another’s authority.  Independence and control are, therefore, different sides of the same coin; assessing whether a person acts independently of another involves considering the level of control or authority that other person has over it or its relevant functions/resources.  The difficulty is in deciding precisely what level of control or dependence suffices in any given case for one entity/its resources to be regarded as permanently available to the other and what links or factors are to be taken into account in making that assessment. 

522.       As noted, we do not consider the Advocate General’s formulation in Welmory, of control over the relevant resources comparable to that of an owner, to be out of kilter with the other authorities.  We also do not consider that, in principle, the required level of control is out of kilter with DFDS.  In that case, the English subsidiary was held essentially to have no independent life from or, in other words, to be under the control of, its parent to the extent that it operated as though it were merely a part of the parent, as its auxiliary organ.  The more material difference is in relation to the links taken into account.

523.       In Welmory, in rejecting the approach that one entity could be a FE of another, the Advocate General by implication seemed to reject that share ownership or common control was relevant to the permanence test (framed in terms of control as she saw it).  The question was what direct rights the entity had over the resources themselves such as under a lease of assets.   

524.       In DFDS, on the other hand, the courts said that the fact the English subsidiary was wholly owned by the Danish parent indicated dependence on the parent. However, that alone was not determinative; they also pointed to the contractual constraints on the English company’s business as set out above.  On the facts in that case those constraints added to and reinforced the dependence indicated by share ownership.  It seems that, equally, ownership or some lesser link (such as common control) is not conclusive if the other facts, such as the contractual arrangements, clearly indicate that the businesses were in fact operated independently notwithstanding the relationship between the parties. 

525.       Both approaches, therefore, require consideration of the nature of the contractual arrangements between the parties as regards the degree of control conferred over the underlying resources/functions or correspondingly the degree of dependence or lack of independence of the relevant entity/its resources.  In that context, the types of factors taken into account by the Advocate General in DFDS as regards the ability to make management decisions and to take financial risk seem equally relevant to the Welmory control test. 

526.       In DFDS the Advocate General also pointed to the fact that the English subsidiary was not able to act for other parties without the consent of its parent as a relevant factor indicating it was not independent (although the CJEU did not expressly refer to that).  That may well not be relevant on a Welmory approach but in any event it is clear that the appellant in this case was able to and did undertake broking activities for other insurers.

Conclusion on permanence test

527.       For all the reasons set out in full below, we have concluded that, on the facts of this case, the appellant’s resources were not available to Advantage with a sufficient degree of permanence for them to constitute its FE on any view of the applicable case law.  Whatever view of the FE test is taken, it is clearly not envisaged that the resources of an entity comprise a FE of another legal (albeit related) entity as a result of the provision of services under commercially agreed contractual arrangements where, in fact, each entity operates a separate business with its own commercial imperatives and financial risk taking. 

Overview of conclusions

528.       To re-cap, as set out in full in the facts section, under the contractual arrangements, in effect, the totality of the functions required to operate an insurance business were split between the two entities.  Advantage acted as the insurer or underwriter which entered into policies with customers (through the appellant) and met claims made by the insured customers.  As insurer it decided what risks to insure for what risk price, set the reserving philosophy and claims handling guidelines and dealt with large loss claims, reinsurance, regulation and accounting, investment of funds received, the making of actuarial reserves and auditing.  It is clear that Advantage managed its own business and was the decision maker on all of these functions albeit that, as regards underwriting and claims handling, it did so with support provided by the appellant.

529.       The appellant dealt with the customer facing side of the business, in selling insurance policies on behalf of Advantage, and handling subsequent claims within the claims guidelines as regards small claims and otherwise subject to direction by Advantage.  It also provided technical and analytical support services to enable Advantage to carry out its underwriting function.  It received recompense for this in the form of its commission which it set itself in fixing the gross premium and, from 2012 onwards, performance related commission relating to claims handling and other targets.  The scope of these functions for Advantage, was defined by the parameters set out in the services agreement and related documents (such as the claims handling guidelines and manuals).  Outside of its relationship with Advantage the appellant also acted as broker or intermediary for other insurers although the majority of its motor insurance business was for Advantage and it had a more extensive role for Advantage. 

530.       The parties were linked beyond the contractual arrangements in the sense that they were under the ownership, for part of the Period, of largely common shareholders and, for the remainder, of a common corporate parent.  Although they were related companies the evidence was that the services agreements were negotiated on an arm’s length basis.  We accept that, on that basis, the division of functions between the two businesses, as made through the contractual provisions, was set commercially as though the parties were independent of each other.

531.       This separation in the two businesses stemmed from the decision to split the underwriting and marketing/retail arms of the business to avoid conflicts and increase profitability for each of the individual businesses (see [76] to [82]).  Mr Charlton said that Advantage could have obtained all of the services provided by the appellant from elsewhere (and we note Advantage had previously sub-contracted to others) but it was a commercial decision or the “best use of capital” to use the appellant (see [162] to [168]).  Mr Lee said that Advantage chose to outsource the data analytics function to the appellant because Advantage thought that it was the “best” (see [259]).  The evidence was that ultimately Advantage could have simply used a different service provider if it was not satisfied with the appellant.

532.       The parties were operating different parts of what could instead be operated as a single business and, inevitably, there were close economic, commercial and operational links between the two businesses.  The two business models were inextricably linked and interdependent for their individual success as they operated together, as related companies, with a common goal, to sell and provide insurance services to UK consumers, at a profit for each of them individually.

533.       It seems to us that the essential difference between the parties was that:

(1) In reliance on DFDS, HMRC, in effect, viewed the overall operations as a single insurance business or economic activity of which, under the contractual arrangements, Advantage, as the insurer and, therefore, the ultimate decision maker or directing authority, retained control on an on-going basis.

(2)  The appellant considered that, unlike in DFDS, the overall operations in this case were successfully split into two businesses which were operated separately according to their own commercial imperatives.

534.       In other words HMRC’s argument was that, through the contractual arrangements, Advantage retained sufficient control of the appellant’s resources on an on-going basis for them to comprise its FE, as the ultimate decision maker on the underwriting side of the business, which approved the parameters within which the appellant operated and which monitored and audited the appellant’s functions within those parameters.  In support of their position, HMRC pointed to the facts that Advantage decided what to insure for what price, it set the policy wording and ultimately retained control over claims handling, directly, as regards large loss claims, and, indirectly, as regards auditing of small loss claims.  The close relationship between the parties and their common or largely common ownership, in their view, added to or bolstered their conclusion. 

535.       In the appellant’s view, the effect of the commercially agreed contractual arrangements was that the appellant and Advantage had distinct functions which, notwithstanding the close degree of co-operation and exchange of information between them, they operated as separate businesses with their own commercial aims and risk taking.  It sufficed to demonstrate that Advantage did not have the required level of control over the appellant’s resources (or that the appellant was acting independently of Advantage) that, within the commercially agreed parameters, the appellant operated its functions autonomously in accordance with its own commercial objectives.  The fact that the two businesses were mutually dependent on each other for their individual success did not detract from that. 

536.       Essentially, we agree with the appellant’s position.  As set out above, it is clear from the decision of the CJEU in Welmory that the fact that parties operate in close-cooperation as part of a single economic activity is not of itself determinative of whether one forms a FE of the other.  Whilst common ownership may be a factor to be taken into account, on the basis of DFDS, as indicating dependence, it is also not of itself conclusive. 

537.       Therefore, the fact that the two related entities operated essentially as different parts of what could be described as a single economic whole does not of itself, without more, lead to the conclusion that the human and technical resources of the appellant, as the service provider, formed a FE of Advantage.  Under both the approach in Welmory and that in DFDS, a closer analysis is required according to the contractual arrangements between the parties and how they operated in practice. 

538.       Where a business delegates or outsources functions which it could otherwise carry out itself, on a basis which, as we accept, was commercially agreed as though the parties were independent, the very fact that the service provider acts within the commercially agreed parameters does not mean, as HMRC seemed to suggest, that the permanence requirement is satisfied.  In other words, we cannot see that, in such circumstances, the setting of the scope of the service provider’s functions itself necessarily amounts to the required retention of control over the relevant resources or demonstrates the required lack of independence.  Again a closer analysis is required of precisely what level of control the business retains as a result of the setting of the parameters and of the functions carried out within those parameters.  In this case, we consider that the answer is clear; the appellant carried out the relevant functions independently of Advantage as a separate commercial enterprise.

539.       On a closer analysis, as the appellant argued, it is clear that, notwithstanding the close relationship between the parties and the linked nature of the two parties’ activities, the appellant acted essentially as an independent service provider to Advantage.  It operated that service business, alongside its more limited business for other insurers, separately and independently from Advantage for its own commercial aims and subject to taking its own financial risk.  The nature of the business, as necessarily linked with and co-dependent on the fortunes of Advantage, does not, given how the business was actually run, lead to any contrary conclusion.

540.       The appellant and Advantage was each run by its own board according to its own commercial aims with its own profit targets and budgets (see, in particular, [145], [152] to [155] and [208] to [224]).  As set out in further detail below, decisions, including as to what level of commission to charge, were taken by the appellant independently with a view to its own interests albeit they may be informed by exchanges of information on Advantage’s plans and that the impact on Advantage was to some extent taken into account.  The appellant took its own financial risk, set its own staff levels and had autonomy as regards the day to day running of the functions it was engaged to undertake.  Advantage monitored and audited the appellant’s function with a view to ensuring that it was providing a satisfactory level of service in each of the areas in which it operated. 

541.       The parties dealt with each other, as regards the performance of the services, as though they were independent parties albeit that they had the advantage of a deeper understanding of the other party’s business than might be expected as regards wholly unrelated parties.  The appellant was free to and did act for other insurers and operated its own business initiatives (such as the counter fraud initiative (see [388] to [392]).  The separation in the roles of the two parties is clear as regards each area in which they operated.

 

Insurance broking and underwriting

542.       As regards the insurance product, Advantage, as the insurer, decided what risks or classes of business to insure for what risk price.  It, therefore, set the net premium, the acceptance criteria and approved the policy wording.  The appellant sold policies to customers through a largely automated system, in the vast majority of cases within criteria pre-set by Advantage but otherwise without interference from Advantage.  It added on to the net premium its own commission, which it set entirely within its own discretion.  It also obtained value from “add-ons” which had nothing to do with Advantage.  Mr Charlton said this was typical of how an insurance/broker relationship works in the market. (See, in particular, [128] to [141] and [156] to [159]). 

543.       The appellant operated as a broker for the other panel members in essentially the same way as it did for Advantage (see [141]).  The sale system did not prefer any insurer over another when selecting the best net premium quote (and the appellant provided quotes to other websites based on the most competitive quote received) (see [138]).  Mr Charlton explained that, accordingly, if a potential customer fell within Advantage’s acceptance criteria, the appellant could nevertheless place the business with another insurer if that business also fell within that insurer’s criteria. It just so happened that Advantage provided the best price in a number of situations (see [162]).  Ms Johnson noted that, having obtained quotes from the insurers, the appellant had a great deal of flexibility as to what customers it did business with (see [284]).

544.       The net premium, which was a pricing of risk, was set by Advantage with a view to achieving its own profit target based on the target loss ratio (the ratio of the costs of claims to premiums) and gross written premium (see [147] to [148]).  The gross premium, in the sense of the final price due from the customer was, as noted, set by the appellant in setting its own commission. 

545.       The appellant took its own commercial and financial risk in setting the gross premium at the right level and in that it also received commission from 2012 onwards on a performance related basis.  The appellant had to pay Advantage the net premium regardless of the final gross premium and even where an amount that was due from a defaulting customer was not recovered (for the length of time that the policy was “live”). 

546.       In setting the gross premium the appellant “faced outwards towards the market”, and assessed how much value it could gain from a customer in the light of market conditions, having regard not only to the level of commission it could earn on the sale of the policy but also to other potential income streams.  It was for the appellant to determine the amount although it took into account the effect on the insurer in a number of respects, including as regards the effect on the mix of business (as that may impact on conditions under reinsurance arrangements) and to some extent the impact on its loss ratio.  It is also provided insurers with information on the impact of changes in the market and areas where they could improve their loss ratio/net premium pricing.  It is clear, however, that the appellant had regard to the insurers’ interests from its own commercial perspective, with a view to maintaining the business relationship with them and with a view to doing whatever it could to obtain competitive net premium quotes from insurers so that it could maximise its own broker value. The appellant performed more detailed analysis for and exchanged a greater level of information with Advantage than for other insurers because it was the insurer for which it did the most business.  (See, in particular, [156] to [159 and [265] to [293].)

547.       Changes to the net premium, acceptance criteria and policy conditions were the way Advantage could seek to meet its loss ratio and gross written premium targets.  It kept the targets constantly under review through the monitoring and data analysis function provided by the appellant.  The appellant used this data to make proposals for change in pricing and risk criteria. In making pricing recommendations to Advantage, the appellant’s team were not considering commission but primarily loss ratio (so covering claims costs) and only secondarily volume or total premium.  (See [145] to [147], [149] to [151] and [225] to [249].)

548.        Whilst the appellant made proposals for changes to the net premium and risk criteria based on this analysis, all decisions were made by Advantage, as the underwriter (see [130]).  Proposals were regularly challenged by the experienced personnel at Advantage.  50% of proposals on net premium proposals were rejected by Advantage (see [250] to [255]).  It was Advantage who made the risk selection. Advantage’s personnel looked at proposals from “their different lens”, as to the selection of the risks that they wanted to write to maximise the value that they got out of the capital that they were managing and how changes could affect its reinsurance and compliance requirements (see [256] to [259]).

549.       We accept that, on the evidence, in particular, as set out at [265] to [321], the net premium and gross premium were set independently by Advantage and the appellant respectively albeit that, due to the impact changes in rates by one party could have on the other and on the overall position in the market, there was an exchange of information on each other’s pricing plans and a degree of collaboration.  It is inherent in the nature of a broker/insurer relationship that both parties have a shared interest in the two components of the gross premium.  Whilst there were sometimes tensions in the parties’ respective pricing positions, in effect the shared interest meant it was usually possible to resolve these by agreement. 

550.       We do not think that the fact that it was often possible to resolve potential conflicts on pricing through agreement indicates that the parties were not acting commercially in their own individual interests. It was simply the case that, due to the interlinked nature of their activities, each party was to some extent dependent on the success of the other party for its own success.  In that context, in many respects the interests of the two businesses were aligned or, to the extent they were not, compromise by one or other party was commercially desirable in taking a longer term position for a sustainable business model.  For example, Ms Johnson noted that it was in Advantage’s interests to have a profitable and successful broker and part of the key profit stream for a broker was commission (although an insurer may query the position if volume was particularly low).  She said that there were “healthy tensions” between the appellant’s desire to maximise commission and Advantage’s need to make its profit but, in cases of conflict, the appellant leaned towards supporting Advantage’s target loss ratio as to do otherwise did not make for a sustainable business model or, as she put it, was “short-termist” (see [294] to [298]).  Clearly if a net premium change had an adverse effect on the appellant such that it was not able to quote, it was in Advantage’s interests to consider that impact as the appellant was its distribution channel (see [291] and [292]). 

551.       It was clear, however, that if the commercial needs of one of the parties justified it that party would act in its own interests.  Mr Godfrey said in effect that if its own commercial needs required it Advantage would alter the net premium even if that was contrary to the appellant’s interests (see [308]).  Ms Johnson gave an example of a case where the appellant had, of its own initiative, changed the gross premium temporarily to reduce volume (as due to a staffing shortage the call centre could not deal with the number of calls) (see [271]).  Mr Lee said that in good market conditions if Advantage saw the appellant was earning large commissions it may decide in effect to take a share of that (by increasing its net premium) not because it needed to but simply because it wanted a share (see [299] to [301]).

552.       Mr Godfrey said Advantage would have taken the gross written premium into account in its budgeting and pricing plans in the same way whether it was using the appellant as its broker or a party such as the AA or Budget; it had to take that into account to know what capital it needed moving forwards.  In his view the degree of collaboration was the same as might take place between Advantage and a third party broker or the appellant and a third party insurer (see [314] and [319]). 

553.       There were also market constraints on the pricing in that, as Mr Gumbrell and Mr Lee noted, if the appellant increased its commissions beyond competitors to a point where Advantage was not able to undertake as much risk as it had appetite for, there was nothing to stop Advantage going back onto other panels like they used to with Kwik-Fit or the AA (for example, see [302]). 

Claims handling

554.       The evidence on claims handling is set out at [169] to [172] and [322] to [387].  We accept that, on the basis of the evidence, the appellant handled small claims on a day to day basis, within the claims handling guidelines and the reserving philosophy set by Advantage, but otherwise without interference by Advantage.  Advantage did not manage its day-to day operations.  As Mr Charlton said, although Advantage tried to control the work flow, it did not have any influence over the appellant’s staffing, how the departments were organised, and how it achieved what it had to achieve within the guidelines.  The guidelines were just that; they were not instructions (see [343] and [344]).  There was little input from Advantage, save in reviewing the handling of the claims and in challenging the appellant’s decision making as part of the audit process.  If Advantage noticed a problem during the audit, it raised the issue with the appellant and it was for the appellant’s management to deal with the issue as they saw fit. 

555.       By contrast large loss claims were reported to Advantage and it was heavily and proactively involved in dealing with them, largely at the LLC, but also through on-going oversight.  The appellant made proposals for reserving and action but it was for Advantage to make the decisions and it was active in challenging proposals.  It set movements in reserves at the LLC meeting, considered whether any large loss claims needed to be reported to its reinsurers (and dealt with the reporting) and dealt with the timing of release of repudiated claims.  The appellant implemented the decisions once made subject to Advantage monitoring the handling of the claim.  Large losses were particularly of concern to the underwriter due to the potential impact on capital and its reinsurance position due to the amount involved.

556.       Mr Lee said that claims handling capability, counter risk and risk selection was “massively important” for Advantage.  We can see that, as the underwriter, Advantage was concerned with claims as its business involved constantly balancing the costs of claims against premiums with a view to maintaining capital and making a profit.  On the other hand, as Mr Lee noted, the appellant, as a separate business and the party dealing with claims on a day to day basis, had to balance its desire to minimise its own spend against the need to manage the spend for Advantage and to treat customers fairly (see [345] to [348]).

557.       It is commensurate with each party’s roles, therefore, as insurer and service provider, that Advantage set the loss claims parameters and dealt with claims which could have the biggest impact.  Mr Lee said that, accordingly, he obtained Advantage’s approval where a claims handling initiative may lead to increased risks for it as insurer, including what he described as strategic risks. 

Monitoring, audits and validation

558.       Advantage had no day to day oversight of the appellant’s work as set out above.  In effect it monitored the level of service provided by the appellant and factors which could impact on that at the ORM through the various reports it received (see [393] to [402]).  It also carried out extensive auditing of the claims with a view to checking they were handled efficiently and checked pricing changes were made correctly (see [372] to [387]).  We do not regard any of this as indicating that Advantage was exercising control over the appellant’s relevant functions as such.  It was simply ensuring it was satisfied with the levels of service provided by its service provider and seeking to prevent any issues arising to disrupt that service.  If issues were identified they were reported to the appellant and it was for it to sort them out.  As regards audits we note that the reinsurers also wanted to see how matters were being handled by the appellant given the importance of its functions to the insurance business (see [366] to [371]). 

Did the resources comprise what was necessary for the insurance supplies to be made on an independent basis?

559.       In case our view on the permanence aspect of the FE test is wrong, we have also considered whether the resources available to Advantage under the contractual arrangements comprised what is “necessary” or “adequate” for the making of the insurance supplies on an independent basis or, as phrased in the Regulation, provided a suitable structure to enable the provision of the supplies.  As noted, in the appellant’s view, this requirement was not satisfied as functions that are integral and critical to the supply of insurance to UK customers were undertaken by Advantage in Gibraltar using its own human and technical resources.

560.       The appellant pointed to Faarborg-Gelting and Welmory as cases where there was no FE due to the lack of key resources/functions at the relevant location needed to make or receive the relevant supplies.  The appellant considered it must have been the lack of management functions in the first case.  In Welmory the CJEU accepted that, if P did not provide C with certain key components which that company needed in order to run its auction business in Poland, it would not have a FE in Poland.  The components identified were computer servers, software, servicing and the system for concluding contracts with consumers and receiving income from them.  In that case, C would not have had available to it resources which would have enabled it to use and receive in Poland the relevant supplies for its own needs in running its auction business.  We accept that, as the appellant submitted, the corresponding position must apply as regards assessing whether a business has the resources necessary for the making of supplies.

561.       HMRC noted that the authorities refer to a minimum level of resources or a structure adequate in terms of relevant resources for the supplies to be made (as in ARO Lease).  HMRC considered that the cases have established, referring to ARO Lease and DFDS in particular, that all that is required as a minimum is that the establishment has the human and technical resources which enable it to enter into contracts with third parties on behalf of the BE autonomously without the need for those parties to deal with the BE. In their view that was clearly the case here.

562.       In HMRC’s view that was the essential difference leading to the different conclusions in those cases.  There was no FE in ARO Lease primarily due to the lack of a framework in which contracts with customers could be drawn up in Belgium.  The Belgian intermediaries did not have authority themselves to deal with the customers other than to refer them onwards to the Netherlands entity.  On the other hand, there was a FE in DFDS because the English subsidiary was able to deal with customers, without directing customers to the Danish parent.  That was the conclusion notwithstanding that the English subsidiary did not have all the technical and human resources which would enable it to operate as a tour operator in the UK.  For example, it used the parent company’s computer to check reservations and it did not have managers who could take important decisions. 

563.       HMRC also noted that in Planzer the CJEU decided that it sufficed for there to be a FE if there is “at least”, therefore, as a minimum, an office where the business could deal with customers by entering into contracts and making daily management decisions.  In RAL the Advocate General specifically rejected the argument that for there to be a FE the relevant business had to possess itself all the resources which gave the establishment its fixed character.  Finally HMRC noted that the lack of a framework for drawing up contracts and dealing with customers’ income was also important in Welmory. 

564.       HMRC queried how the FE test could operate on the appellant’s interpretation where the issue is whether a branch of an insurance company is a FE.  On the appellant’s argument a branch could not be a FE because of necessity it would not have all the resources needed to supply insurance.  The head office is the part of the operation that would have to meet relevant regulatory requirements, raise capital and supervise the underwriting risk.  However, it is clear that a branch can be a FE of an insurance company as demonstrated by the decision in HM Revenue & Customs v Zurich Insurance Company [2007] EWCA Civ 218.  In that case the Court of Appeal upheld the decision that supplies of consultancy services made by PwC relating to the installation of software under a framework agreement with a Swiss insurance company were provided at the company’s FE in the UK (where the software was to be used). 

 

 

Conclusion on whether the resources comprised what was necessary for the supplies to be made on an independent basis

565.       We accept that for an entity to have a FE in a particular location does not require that the relevant human and technical resources permanently available to it in that location provide absolutely everything needed to put the business in the position where it can make the relevant supplies from that location.   It cannot be the case that there is a FE only if the relevant business devolves the entirety of its operation to the place in question.  A business may have a FE notwithstanding that it draws on certain limited functions of the BE, such as the computer facilities for checking reservations in DFDS, or if it buys in or outsources certain of the relevant functions, as the Advocate General said in RAL.  It seems to us that it is a question of fact and degree in assessing in each case what resources are “necessary” for the making of the supplies in the sense, as the Advocate General put it in RAL, of those which are “directly involved” in the supplies. 

566.       In the context of the particular businesses in ARO Lease and Welmory, the presence or absence of resources in the relevant location providing the framework for the business to contract with customers was a material factor.  In ARO Lease that was the conclusion in the context of supplies which, as the court said, consisted principally in negotiating, drawing up, signing and administering the relevant car leasing agreements and in making the vehicles concerned physically available to customers. 

567.       We do not understand the courts in these cases to be laying down a principle that the ability of an establishment to conclude contracts or deal with the “customer facing” side of the business, without reference to the BE, is a minimum threshold above which there is necessarily a FE (assuming the other requirements are satisfied).  The decisions in these cases do not preclude the possibility that there may be other resources/functions which may be “necessary” for the making of the relevant supplies “on an independent basis” on the particular facts of a case.  Indeed there were other considerations such as, in Welmory, the technical resources needed to run an online auction, such as servers and software and, in that case and, in ARO Lease, the ability to make management decisions. 

568.       As the Advocate General noted in RAL, what constitutes a necessary framework for the making of supplies will vary according to the particular context.  It is a question of focussing on the particular supplies and assessing what was directly involved in enabling the business to make those supplies.  Whilst the ability to conclude contracts without reference to the BE may suffice for this requirement in the context of a car leasing business, it does not follow necessarily that it is sufficient in the context of other businesses. 

569.       We do not regard DFDS as detracting from this.  In that case, the focus was on the subsidiary’s dependence on the parent such that it was an auxiliary organ only.  The point was that the English subsidiary was so constrained that it operated as though it were part of the parent, or in conjunction with but subordinate to the parent, as though the two companies were a single unit.  The CJEU considered otherwise that on the facts of that case there were sufficient resources in the UK for the establishment there to satisfy the minimum requirement due, in outline, to the presence of large numbers of staff in the UK and as contracts with customers were initiated in the UK.  The courts were not laying down as a principle that in all cases there must be a FE where the resources in the UK initiate or conclude the contract with customers. 

570.       In this case the supplies in question are the conclusion and performance of contracts of insurance with customers in the UK.  In return for the premium the customer obtains insurance cover for a motor vehicle for a specified period under which it can, subject to the specified conditions, make a claim for the insurer to cover costs and damages in certain eventualities. 

571.       It is not disputed that the appellant dealt with almost the entirety of the “customer facing” side of the business, in the sense that it sold the insurance policies on behalf of Advantage and administered the policies by dealing with customer queries, complaints and the day to day handling of any claims the customer later made (albeit, as regards large loss claims, under direction from Advantage). 

572.       In a business involving on-going supplies of insurance services, clearly the supplies could not be made without the resources to sell and administer the policies.  However, in our view, equally, such supplies could not be made without the resources which provide the ability actually to underwrite the risk or, in other words, to provide the insurance cover.  That is essential to the making of insurance supplies in the same way as technical resources, such as servers and software, may be held to be essential for the running of an online auction site. 

573.       In our view that is the key factor.  From the evidence set out above, it was clearly Advantage which, as the insurer, using its own staff and resources in Gibraltar, provided the insurance product and the insurance cover.  It decided what risks to insure for what risk price and, having written the policy, assumed liability and paid out the funds on settlement of a claim under the policy.  For all the reasons set out above, that it received input and analysis from the appellant to enable it to make its underwriting and related decisions does not detract from the fact that it was Advantage which managed its own underwriting business.  The appellant sold the insurance, on behalf of Advantage, under a largely automated process.  In the vast majority of cases sales were made within risk acceptance criteria and for a risk price or premium set by Advantage (albeit the appellant added its commission).  In administering the policies as regards subsequent claims the appellant acted within the guidelines or, as regards large loss claims, subject to instruction from Advantage.

574.       On that basis, whilst there were many other related functions which Advantage undertook using its own resources in Gibraltar, it is not necessary to look any further.  The appellant’s resources did not provide a key element of what was required for the supply of insurance services, namely, the ability to decide what/who to cover for what risk price.  Without that essential function the appellant could not have sold insurance cover to UK customers at all. There would not have been a product for it to sell.

Discussion - were the services supplied to a UK FE of Advantage?

575.       The next question is whether, although we have found there was no UK FE through which Advantage made supplies of insurance, the resources of the appellant nevertheless comprised a FE at which the services supplied by the appellant were received by Advantage.  We cannot see how that could be the case as (a) for all the reasons set out above, in this context also, the appellant’s resources did not constitute an establishment characterised by a sufficient degree of permanence to be a FE and (b) they did not provide a suitable structure to enable it to receive and use the services supplied to it for its own needs, given that Advantage made the insurance supplies from its BE in Gibraltar. 

576.       As the Advocate General said in Welmory, at [43], it must be “doubtful whether as a rule every structure which, in terms of its human and technical resources, is able to use services for its own needs would not indeed at least have the possibility of supplying services itself”.  On our findings there could not be the possibility that any any FE in the UK was making insurance supplies; they were made by Advantage from its BE in Gibraltar. 

577.       If, contrary to our view, HMRC’s analysis is correct, the position would be that the appellant/its permanently available resources comprised a UK FE, as regards supplies made by Advantage, due to the customer facing role undertaken through those resources. We can see that it is arguable that, on that basis, it follows that the UK FE of Advantage was capable of receiving and using for its own needs the services relating to those customer facing functions (being, broadly, those used to sell the policies to customers and to deal with day to day claims handlings and customer complaints).  (That is subject to the argument that actually there is no separate supply of services at all in those circumstances (see [582] to [585] below).

578.       However, on HMRC’s own analysis, we cannot see that it follows that the resources comprising any such UK establishment enabled Advantage to receive and use for its own needs at that establishment other services, in particular, the underwriting support services or those relating to large loss claims.  The underwriting support services enabled Advantage’s personnel in Gibraltar to monitor its targets and decide whether to make changes to its net premium or risk acceptance criteria.  The appellant’s input on large loss claims was used by the personnel in Gibraltar to decide how such claims were to be handled, what reserves should be made and to provide relevant information to the reinsurers.  It is not disputed that these functions were carried out by Advantage’s own staff in Gibraltar.  In our view, as a matter of commercial and economic reality, therefore, any such UK FE would only be capable of receiving and using the broking, small claims handling and related client handling functions.  The appellant’s resources did not provide what was needed for Advantage to use the other underwriting support and claims handling services for its own needs as insurer.

579.       If the different services provided by the appellant comprised separate supplies, the logical outcome would seem to be that (a) the resources comprised a UK FE as regards the client handling services which accordingly were supplied to that UK FE and (b) the remainder were supplied to the BE in Gibraltar.  The input tax in dispute could be apportioned between the two sets of supplies. 

580.       However, the parties appear to have been proceeding on the basis there was a single composite supply of services, as accords with the fact that, in the services agreements, there was no separate attribution of value to the different elements.  On that basis, it is difficult to see that there can be anything other than an “all or nothing” approach.  It seems to us that the conclusion, assuming there was a single supply of all the services, is that any such establishment did not constitute a UK FE of Advantage at which the services were supplied, as the relevant resources in the UK were plainly not adequate or suitable to enable it to receive and use for its own needs the totality of the services, as single sets of supplies.

581.       If, contrary to our view, the fact that the resources enabled some part of the supplies to be received and used in the UK suffices for there to be a UK FE, the allocation of the supplies to the UK FE or to the BE would have to be dealt with again it seems on an “all or nothing basis” under the rationality “tie breaker” test as discussed below.

582.       Finally, there is further complexity in that we see a conceptual problem with the premise underpinning HMRC’s analysis.  It involves the proposition that the appellant was both acting (a) as a FE of Advantage, through which Advantage made insurance supplies and at which it received the services and (b) separately, in a different capacity than as FE of Advantage, in providing those services to that FE. 

583.       In Welmory the Advocate General accepted that P’s resources could be a FE of C whilst at the same time P was, in a different capacity, a service provider to that FE.  However, she recognised that if the relevant resources of P, as service provider, and of the establishment of C are “virtually the same, it may be questioned whether there is a supply of a service to another taxable person at all”. 

584.       In this case, we can see that the appellant could be said to provide some of the services (such as the underwriting support services) to a UK FE of Advantage as service provider in a different capacity to it acting as such a FE.  However, the resources used in providing the broking, small claims handling and related customer services to Advantage are the very resources which HMRC argued comprised a FE as a result of the functions carried out through them.  As the Advocate General said, where the resources comprising the FE and used in making the services are the very same, there cannot really be a supply of a service to another person at all.  The outcome of HMRC’s analysis in fact appears to be, therefore, that, at least as regards those services, the UK FE of Advantage supplied them itself as FE, to itself as FE. 

585.       It follows that there is in fact no supply of those services by or to another entity at all.  Advantage would simply be making its supplies of insurance using the relevant resources which carry out the functions giving rise to the FE.  We note again the difficulty of splitting out the services in effect into different supplies/elements but we cannot see how an entity can provide services to itself (in the absence of any deeming provision) so that such an allocation may be necessary.  It would seem to follow that in principle the input tax in dispute would need similarly to be apportioned. We note that, however, this position is far from the commercial and economic reality of this case.  We have not considered it further given that, on our view, the issue does not arise in any event.

Discussion - choosing between the place of BE and FE

586.       Finally we have considered the rationality test in case we are wrong that there was in fact no FE so that the issue of choosing between the BE in Gibraltar and a UK FE does not arise.  It is assumed for the purposes of this section, therefore, that there was a UK FE for the reasons submitted by HMRC.

587.       HMRC continued with their two stage analysis.  First, they argued that it is justifiable to depart from the place of the BE, as the place where Advantage supplied insurance to UK customers, for similar reasons to those set out in DFDS and RAL.  In DFDS the location of the UK FE was preferred on the basis that (a) under general principles VAT should be charged at the place of consumption and where the services were marketed (being the UK) and (b) taxing by reference to the BE would potentially lead to distortion of competition (as businesses could choose to base themselves in member states which exempted the supplies in question) (see [430] to [438] above).  In RAL taxing by reference to the BE in Guernsey would have meant that the supplies would be outside the scope of EU VAT.  The risk was that taxing by reference to the BE would encourage companies to establish their businesses outside the EU, while continuing to supply their services in the EU through FEs located there to consumers residing there (see [455] and [456] above). 

588.       HMRC said that similarly in this case the place of consumption and where the services were marketed was clearly the UK.  Accordingly, exactly the same concerns arise as in those cases in looking to Gibraltar, as the place of supply/belonging, as regards the place of supply of the insurance services. 

589.       Secondly, HMRC continued that, once it is established that Advantage made its supplies through the UK FE, it follows it would also be irrational to treat the BE as the place of supply/belonging as regards the services provided by the appellant.  In HMRC’s view the actual economic situation was that the services were “predominantly” used or consumed at the UK FE to market and sell insurance to UK customers.  Moreover, to tax by reference to the BE would, therefore, enable the appellant to recover input tax on exempt supplies made in the UK thereby leading to potential distortion of competition.

590.       HMRC took further support for this view from the High Court decision in the Zurich case (referring, in particular, to [40], [49] to [51] as approved by the Court of Appeal ([25], [45], [46], [54] and [55] and [101]).  HMRC noted that, at [40], Mr Justice Park acknowledged that the services in a sense benefited the Swiss head office as well as the UK branch as it was the head office which contracted for their performance.  He held, however, that the actual provision of the services to the branch (as the services related to the replacement of the branch’s computing system) “far outweighs in importance” that the contract which the consultants thereby performed in the UK was made with the head office.  He said that “in reality”, the services were supplied to the UK FE.   

591.       As regards the rationality test, he said, at [49], the concern on non-taxation, was that if services are “consumed” in the EU they should be subject to a VAT charge; in his view, these services were consumed in the UK “to a much greater extent” than they were in Switzerland.  He continued, at [50], that the UK branch competed with other insurance enterprises doing business in the UK, most of which were UK companies.  Such competitors would have to pay VAT on such consultancy services but would be able to recover only a small proportion of the VAT.  He concluded, therefore, that for the UK FE to be able to replace its computerised system without a VAT charge was a “manifest distortion of competition”.

592.       HMRC also referred to the comments of Moses J, as he then was, in Customs and Excise Commissioners v Chinese Channel (Hong Kong) Ltd [1998] STC 347 where he said, at [354], that it was important “to consider the significance of [the] activities and the part they play in their contribution to the service supplied”.  In that case he held that the tribunal was entitled to find that the main contribution in broadcasting a TV channel was made by the Hong Kong establishment, since it selected programmes, made the relevant contracts for the rights to programmes and made arrangements for transmission.  The UK establishment played a lesser role in providing administrative and accounting facilities, as well as some production and editorial functions.  HMRC said that, in this case, as Mr Kendall said, the most significant activities were those of the effecting and carrying out of the insurance contracts which was carried out for Advantage by the appellant.

593.       HMRC concluded that, in such circumstances, it would be fiscally irrational to allow input tax recovery to a UK entity supplying insurance services to an insurer in Gibraltar UK supplier making similar supplies to a UK insurer could not recover such input tax.  Such a result would be contrary to recital 7 of the Directive which states that: “The common system of VAT should…result in neutrality in competition, such that within the territory of each Member State similar goods and services bear the same tax burden, whatever the length of the production and distribution chain”.

594.       The appellant’s primary contention was that the BE was to be preferred as the place of supply/belonging as regards the supplies of services made to Advantage whether or not Advantage was regarded as making supplies of insurance through a UK FE.   However, the appellant considered that, in any event, it was not correct that the UK FE was to be preferred as regards the place of supply of the insurance services. 

595.       The appellant noted that the fact that contracts were concluded in the UK and that the consumers were in the UK had no bearing as it considered was clear from Zurich and Welmory (at [64]) respectively.  It said that the conclusion in DFDS cannot be taken to have general application.  The Advocate General relied on the opinion of the Advocate General in Berkholz but the CJEU in that case did not endorse that view.  He also referred to the BE and FE as though they were each main criterion. However, whilst that may be justified by reference to TOMS, the reinforcement of the primacy to be accorded to the BE in Welmory make it clear that is not correct as regards the general place of supply rule.

596.       The appellant noted that in ARO Lease the CJEU held that consumption was not a meaningful way of determining the place of supply of transport given the difficulty in determining where the vehicle is actually used.  In the appellant’s view, there is the same difficulty here in assessing where the insurance supplied by Advantage was used as that depends on where the insured vehicle was used.  The Advocate General in that case recognised, at [27], that “the mere fact that a member state may suffer financially as a consequence” of adopting the BE as the place of supply “is not relevant”.  On the contrary the alternative “exceptional [FE] point of reference may only be applied” where the relevant conditions are satisfied and “if, in that eventuality, the application of the [BE] would be unreasonable”. 

597.       The appellant considered that distortion of competition concerns are not relevant to this test.  The fact that companies carrying out similar transactions in the UK might be exposed to a higher tax burden is not of itself an “irrational” result.  The concerns raised are simply the inevitable consequence of a non-harmonised VAT system coupled with the ability of a company to exercise its EU law right to freedom of establishment.  In this case, as a political decision, Gibraltar is not within the EU for VAT purposes but is inside the EU for financial services purposes.  Advantage exercised its right to supply insurance to customers in the UK from Gibraltar relying upon its EU passport rights.  There has been no suggestion that the decision by Advantage to locate its business in Gibraltar is driven by fiscal considerations.  Absent abuse, which is not alleged here, the tax consequences must follow the business structure that has been adopted.

598.       The appellant argued that is supported by CJEU decision in Case C277/09 The Commissioners for Her Majesty’s Revenue and Customs v RBS Deutschland Holdings GmbH, [2011] STC 345 (at [53]) where, as regards an abuse argument, the court noted that taxable persons are generally free to choose the organisational structures and the form of transactions which they consider to be most appropriate for their economic activities and for the purposes of limiting their tax burdens.  The appellant also referred to the decision of Moses J (as he then was) in the Chinese Channel where he said, at [356 (e-f)], that:  “It is by no means clear that…[non-taxation] is a relevant consideration where non-taxation only occurs outside the [EU].....”. 

599.       We note, however, that, as HMRC pointed out, that comment was not part of the binding decision in that case.  HMRC also cast doubt on the applicability of the RBS decision on the basis it was an abuse of rights case which was not concerned with place of supply/FE issues as such. The appellant responded that the principle emerging from RBS is clearly of general application, namely, that, in an unharmonised system, tax loss is likely to occur but unless that loss arises as a product of an abusive practice, it is simply an outcome that has to be lived with. 

600.       The appellant continued that it could hardly be an economically irrational conclusion to treat the supplies of services as made to the BE in Gibraltar given that, although the services may be capable of being used in part by Advantage at a UK FE, they were, as a matter of economic reality, also “used” by Advantage at its BE in Gibraltar.  The Zurich case has no bearing on this; the point there was that the use of the consultancy services at the office in Switzerland was ancillary or minimal only.  The use in Gibraltar in this case is substantial and significant.

601.       The appellant asserted that this result is entirely consistent with the case law and is mandated by article 22 of the Regulation.  The criteria in the Regulation are directed to economic rationality and not any separate concern as to fiscal irrationality. The same points on fiscal irrationality and tax loss apply as set out above.

602.       Under article 22 it is provided that, if an examination of the nature and use of the service does not enable the FE to be identified, particular attention must be paid to the contract, the order form and the VAT identification number and whether the FE is the entity paying for the service.  Where the FE cannot be determined in accordance with those rules or where services are supplied to a taxable person under a contract covering one or more services “used in an unidentifiable and non-quantifiable manner”, it is legitimate to consider that the services are supplied where the BE is located (see [23] above). 

603.       The appellant thought that it was clear from those criteria that the services were supplied to the BE.  The agreements were made between the appellant and Advantage acting through its BE for the provision of services by the appellant on arm’s length terms (with provision for review and discussion of the required service levels on a monthly basis) in return for commission due from Advantage again acting through its BE.  If that is not accepted, as it is clear that services were at least in part and in a significant way provided to the BE, at best it can be said that the services were used in an unquantifiable or unidentifiable manner.  The appellant noted that there is no division in the service agreements in terms of time and resources between the various services and no allocation of the appellant’s commission between the services.  Moreover the agreements provided no basis for any such allocation according to the extent a particular service, such as claims handling, which was clearly used in both countries, was used in the UK and Gibraltar respectively.  In cases of uncertainty the law is clear; the presumption in favour of the BE is not displaced.

604.       HMRC responded that the applicable provision in the Regulation is in fact article 21 (see [23] above).  That article deals with the situation where there is one BE and one FE in two separate countries and allocates taxing right to the place where the FE is if the supplies are made to the FE.  Article 22 in fact deals with cases where there is more than one FE in different member states.  These rules are designed not to supplant the case law of the CJEU but in fact to apply it; in this case, it is article 21 which encapsulates the relevant case law.  Article 22 cannot be applied by analogy to the circumstances of article 21 because these rules are intended to be applied only to the relevant specific circumstances (see the sections in the preamble set out at [20] above).  The appellant countered that was not supported by the wording and plain intention of the provisions.

Conclusion on BE versus FE

605.       On the first issue as to the correct place of supply of the insurance services, as in DFDS and RAL, the place of consumption and that where the services were marketed was where the FE was located, in the UK.  We consider that the place of consumption of insurance services is the place where the insured is located, as the person benefitting from the insurance cover.  That is not the same as the case where the supply comprises making a car physically available for use as a means of transport under a leasing contract.  On that basis, similar concerns arise as in DFDS and RAL that the supplies of insurance would be outside the scope of EU VAT if the BE is used as the place of supply/belonging.  However, the courts in DFDS were seemingly influenced by the particular issues relating to supplies by tour operators and travel agents.  Moreover there must be real doubt that such issues are relevant following the comments in the RBS case.

606.       In any event we do not consider it necessary to reach a definitive view on that issue.  Even if it is correct that Advantage made its supplies of insurance through a UK FE, we do not consider that it necessarily follows that it is justifiable to depart from the BE as the place of supply/belonging as regards the supplies of services made by the appellant to Advantage.  HMRC’s position was that there was a UK FE due to the fact that the appellant’s resources in the UK (which they considered to have the required degree of permanence) dealt with the marketing and sale of insurance and related “customer facing” activities without reference to Advantage.  It was not disputed that Advantage was in fact the insurer/underwriter operating its business in Gibraltar with its own staff and resources located there.  It was not HMRC’s case on the FE issue that that function was provided by the appellant’s resources in the UK. 

607.       On that basis we cannot see that it follows, as HMRC said, that the services provided by the appellant “predominantly” related to the marketing and sale of insurance activities carried out in the UK. Whilst the broking and some of the claims handling services clearly did relate to that function it is equally clear that, in reality, Advantage used the underwriting support and some of the claims handling services in carrying out its underwriting function in Gibraltar.  As set out above, the difficulty is that we have no ready means of determining the real extent or value of the services, on this analysis, used at any UK FE and those used at the BE in Gibraltar.  We agree, however, that, as the appellant submitted, those used in Gibraltar are plainly not ancillary only as was the situation in the Zurich case.  We can see no basis for assuming that they predominantly relate to the “client facing” activities carried out in the UK or that those activities are somehow more important than the others involved in the operation of a motor insurance business (see also our comments at [83] to [91] above).

608.       This seems to be precisely the sort of case where, under the guidance in regulation 22, it is legitimate to look to the BE as the place of supply/belonging on the basis that the services fall into the category of services which are “used in an unidentifiable and non-quantifiable manner” or that this is a case of doubt.  That guidance tells us that, as taxing rights have to be allocated to a single country, in such circumstances it is appropriate to rely on the place where the BE is, as providing a certain criterion.

609.        We do not accept HMRC’s argument that regulation 22 does not apply in these circumstances.  The wording does not suggest it is confined to the situation where there are a number of FEs and a BE as opposed to where there is a single FE in addition to the BE.  Moreover we can see no rationale for any such distinction. The regulation is clearly aimed at providing assistance in determining the place of taxation/supply where there is more than one establishment in question and the establishments are in different locations.

610.       The end result is that the appellant is able to recover all of the input tax in dispute notwithstanding that some of it, on HMRC’s analysis, may be attributable to the broking and claims handling services which, on that analysis, were used by the UK FE of Advantage to make exempt supplies of insurance in the UK.  However, in these circumstances, we cannot see that is sufficient basis to disregard the primacy to be accorded to the BE as the place of belonging/supply.  The rules specifically allow such input tax to be recoverable where it relates to supplies of this kind made to a person who belongs outside the EU.  On the basis of the principle stated in RBS, a business is free to choose where to establish its business, whether with a view to minimising its tax burden or otherwise.  Advantage was established in and was carrying out its business in Gibraltar and the services provided by the appellant were in a very real and substantial sense used by it in operating its business from there. 

Conclusion

611.       For all the reasons set out above, we have concluded that the input tax in dispute is recoverable on the basis that it is attributable to supplies made to Advantage on the basis that it belonged outside the EU (as interpreted in accordance with the relevant EU rules and case law). 

(1) The appellant’s human and technical resources, through which it provided the services to Advantage, did not comprise a FE of Advantage in the UK, whether for the purposes of determining where Advantage made supplies of insurance or where the appellant made the supplies of its services. 

(2) Even if, contrary to our view, those resources comprised a FE in the UK, there is no reason to depart from the location of Advantage’s BE in Gibraltar as the place of belonging/supply in the circumstances of this case. 

612.       Accordingly, the appeal is allowed.

613.       This document contains full findings of fact and reasons for the decision. Any party dissatisfied with this decision has a right to apply for permission to appeal against it pursuant to Rule 39 of the Tribunal Procedure (First-tier Tribunal) (Tax Chamber) Rules 2009. The application must be received by this Tribunal not later than 56 days after this decision is sent to that party.  The parties are referred to “Guidance to accompany a Decision from the First-tier Tribunal (Tax Chamber)” which accompanies and forms part of this decision notice.

HARRIET MORGAN

TRIBUNAL JUDGE

 

RELEASE DATE: 19 JANUARY 2018