xmlns:atom="http://www.w3.org/2005/Atom" xmlns:atom="http://www.w3.org/2005/Atom"

Commission Decision

of 11 December 2007

on the State aid case C 53/06 (ex N 262/05, ex CP 127/04), investment by the city of Amsterdam in a fibre-to-the-home (FttH) network

(notified under document number C(2007) 6072)

(Only the Dutch version is authentic)

(Text with EEA relevance)

(2008/729/EC)

THE COMMISSION OF THE EUROPEAN COMMUNITIES,

Having regard to the Treaty establishing the European Community, and in particular the first subparagraph of Article 88(2) thereof,

Having regard to the Agreement on the European Economic Area, and in particular Article 62(1)(a) thereof,

Having called on interested parties to submit their comments pursuant to the provisions cited above(1) and having regard to their comments,

Whereas:

I. PROCEDURE

(1)In April 2004, the municipality of Amsterdam contacted the Commission regarding the public procurement aspects of the roll-out of a FttH (fibre-to-the-home) telecommunications access network. In addition, the municipality requested the Commission to confirm that the project did not entail State aid within the meaning of Article 87(1) of the EC Treaty. By letter of 22 July 2004, the Commission informed the municipality of Amsterdam that such a confirmation can only be given after a notification of the measure by the Dutch authorities. The Commission requested the Dutch authorities on 23 July 2004 to provide any information necessary for an assessment of the measure under Article 87(1) of the EC Treaty. The Dutch authorities asked for an extension of the deadline in August 2004 which was accepted by the Commission on 7 September 2004.

(2)In September 2004, the Dutch authorities convened with the Commission to present and discuss the plans of the municipality of Amsterdam. On 7 October 2004, the authorities stated in their reply to the Commission’s letter of 23 July 2004 that the project in Amsterdam would be notified in the near future. On 17 May 2005, the Dutch authorities notified the project — the participation of the municipality of Amsterdam in an undertaking carrying out the roll-out and owning this network. The Dutch authorities were seeking confirmation from the Commission that the investment of the municipality of Amsterdam in the legal entity owning the network is in line with the Market Economy Investor Principle (‘MEIP’) and accordingly does not constitute State aid.

(3)Following further information sent by the authorities on 23 June 2005, and a meeting which took place on 28 June 2005 between representatives of the city of Amsterdam and the Commission, the Commission sent on 15 July 2005 a letter containing elements of an explanation of the application of the MEIP together with a second request for information.

(4)The Dutch authorities stated in a letter, registered on 18 November 2005, that there was a delay in the planning and that the municipality of Amsterdam was still working on the setup of the project and the investment conditions. The Dutch authorities declared that they would need more time in order to provide the requested information and asked the Commission to suspend the assessment until all data would be available.

(5)The municipality of Amsterdam informed the Commission by e-mail registered on 23 December 2005 that the City council of Amsterdam had by unanimity decided to invest in the roll-out of the FttH network and stated further that negotiations with BAM/DRAKA (for the construction of the network) and with BBned (for the exploitation of the network) and the negotiations with ING RE and five housing corporations (co-investors) were all on track and would be finalised by January 2006.

(6)By letter registered on 27 December 2005, the Commission received a complaint regarding the project from VECAI (the association of cable operators in the Netherlands, which changed its name in September 2007 into ‘NLKabel’)(2). UPC Nederland BV (a cable operator present notably in Amsterdam, hereinafter: ‘UPC’(3) informed the Commission for the first time about its concerns in March 2005. Both parties — with whom several meetings took place — argue that the participation of the municipality is not in line with the MEIP and constitutes State aid within the meaning of Article 87(1) of the EC Treaty.

(7)On 3 March 2006, the Commission sent a reminder to the Dutch authorities, referring to the statement of the Dutch authorities in November 2005 that further information would be provided once further progress had been made on the setup of the project and which was foreseen for spring 2006. The Commission also reminded the Dutch authorities of the standstill obligation of Article 88(3) of the EC Treaty.

(8)The Dutch authorities replied to the Commission’s letter of 3 March 2006 on 3 April 2006, and sent additional information by letter of 2 May 2006. On 19 May 2006, the Commission’s services sent another letter to the Dutch authorities, referring to the information received from UPC which cast initial doubts on the application of the MEIP and reminded the authorities again of the standstill obligation(4).

(9)The Dutch authorities sent further information in May and June 2006(5). The Commission sent an additional request for information to the Dutch authorities on 24 July 2006, in answer to which the Dutch authorities submitted several batches of information(6). Following an additional request of information of 29 September 2006, the authorities asked on 13 October 2006 for an extension of the deadline to reply.

(10)In the meantime, the roll-out of the network commenced on 12 October 2006(7). In view of these developments, the extension of the deadline requested by the Dutch authorities was refused. The Dutch authorities submitted part of their answers before the deadline of 26 October and a substantial amount of information was also provided on 30 October, 16, 20 and 21 November 2006.

(11)UPC requested the District Court in Amsterdam in a procedure for interim measures to order the municipality of Amsterdam to respect the standstill obligation laid down in Article 88(3) of the EC Treaty and not to continue the project before the Commission has finalised its assessment. The request by UPC was dismissed by the District Court which stated in its judgement of 22 June 2006 that it was not obvious that the municipality’s involvement in the project involved State aid. According to the Court, the mere fact that the municipality has initiated the project is not to be considered as State aid. Moreover, the Court concluded that the initial costs incurred by Amsterdam (studies, etc.) would be reimbursed by the joint-venture Glasvezelnet Amsterdam CV and therefore did not constitute aid.

(12)UPC informed the Commission by letter of 22 September 2006 that it had appealed the decision by the District Court. The municipality of Amsterdam(8) informed the Commission in November 2006 that the Court of Appeal proposed to UPC and the municipality of Amsterdam to await the Commission’s decision before a judgement by the Court would be rendered. Both parties accepted this proposal(9).

(13)By letter dated 20 December 2006, the Commission informed the Netherlands that it had decided to initiate the procedure laid down in Article 88(2) of the EC Treaty in respect of the notified measure.

(14)By letter of 8 January 2007, the Dutch authorities requested access to several documents submitted by UPC mentioned in the opening decision. By letter registered on 13 February 2007, UPC agreed to share the requested documents with the Dutch authorities, which were forwarded by the Commission to the authorities.

(15)On 18 January 2007, the Court of Appeal Amsterdam(10), rejected the appeal brought by UPC Nederland against the municipality of Amsterdam to stop further activities related to the measure. The Court of Appeal found that, although the Commission expressed in its opening decision certain doubts as to whether the MEIP is fulfilled, by providing the requested information to the Commission, the municipality of Amsterdam could still prove that the measure is in line with the MEIP and does not constitute aid. Under these circumstances, the Court of Appeal considered it possible that the Commission would conclude that no aid is involved. The Court also found that the municipality would be hurt if it would have to stop all activities related to GNA, while UPC would not significantly suffer from the preparatory activities undertaken by Amsterdam.

(16)By letter of 5 March 2007, UPC asked the Commission for access to confidential information submitted by the Dutch authorities and to set up a ‘data room’(11) procedure for the measure at hand, in order to be able to assess the Dutch authorities’ claims to substantiate their complaint further. By letter of 26 March 2007, the Commission informed UPC that the State aid procedures do not foresee access for third parties to the requested confidential information or for setting up a ‘data room’ procedure.

(17)The Dutch authorities responded by letter registered on 16 March 2007 to the Commission’s decision to open the formal investigation procedure.

(18)After several exchanges of correspondence between the Commission and the Dutch authorities on confidentiality issues, the public version of the opening decision was sent to the complainants on 23 April 2007 and published in the Official Journal of the European Union (12). The Commission invited interested parties to submit their comments on the measure.

(19)Following the opening of procedure, the Commission received comments from five interested parties, namely from Liberty Global/UPC(13), Com Hem(14), Ono(15), France Telecom(16) and another party which requested anonymity(17). By letter registered on 30 July 2007, the Commission forwarded the third party comments to the Dutch authorities. The authorities submitted their observations by letter dated on 17 September 2007.

(20)The Commission services met with representatives of Liberty Global/UPC on 5 July 2007 and with the Dutch authorities on 5 November 2007. The Dutch authorities submitted additional information on 9 November 2007 and on 12 November 2007.

II. DESCRIPTION OF THE MEASURE

(21)The municipality of Amsterdam decided in 2004 to investigate the possibility of investing in a fibre-to-the-home (FttH) electronic communications access network(18). Following reports, market analyses and preliminary contacts with potential investors, the municipality decided formally in December 2005 on the said investment and the conditions under which such an investment would take place.

(22)The planned FttH network in Amsterdam will serve 37 000 households in Amsterdam in the districts of Zeeburg, Osdorp and Oost-Watergraafsmeer, which comprise together about 10 % of the city of Amsterdam. The municipality of Amsterdam also expressed its long-term ambition to extend the project to other parts of Amsterdam, with up to 400 000 household connections altogether. However, this potential extension is not part of the project under assessment.

The three-layer model

(23)The project in Amsterdam is based on a ‘three-layer model’ as outlined below.

Figure 1

Simplified overview of the Citynet Amsterdam project

(24)The first layer is the so-called ‘passive network infrastructure’, which includes ducts, fibre and street cabinets.

(25)This passive infrastructure is activated by means of telecommunications equipment (the second or ‘active layer’) by a wholesale operator managing and maintaining the network and providing wholesale services to retail operators. The active layer includes the management, control and maintenance systems necessary to operate the network, such as switches, routers or splitters.

(26)The wholesale operator will provide an open, non-discriminatory access to retail operators to offer television, broadband, telephony and other Internet-based services to end customers (third or ‘retail layer’). In order to be able to provide these services, retail service providers will have to invest, inter alia, in equipment, procure content and operate their own service platform (maintenance, customer care, and billing).

The stakeholders of the project

(27)The passive infrastructure will be owned and managed by the partnership Glasvezelnet Amsterdam cv (‘GNA’)(19), whose shareholders are the municipality of Amsterdam, two private investors, namely ING Real Estate (hereinafter: ‘ING RE’)(20) and Reggefiber(21), and five subsidiaries of social housing corporations(22). A cooperation agreement (‘Samenwerkingsovereenkomst’) was signed by the parties on 11 April 2006 and the investment agreement (‘CV-overeenkomst’) was signed on 26 May 2006.

(28)The municipality of Amsterdam invests EUR 6 million, ING RE and Reggefiber invest each EUR 3 million, three social housing corporations invest each EUR 1,5 million and two housing corporations invest each EUR 750 000(23). The total equity investment amounts to EUR 18 million. In addition, GNA acquired loan amounting to EUR […](24) from […](25). The total planned budget for the project amounts to EUR 30 million. As also outlined in Figure 1, the Amsterdam municipality owns one third of the shares of GNA, the two private investors together another one third, while the subsidiaries of the housing corporations own the remaining one third.

(29)The passive infrastructure of GNA will be activated and operated by BBned(26) on the basis of a contract between GNA and BBned of […]. The necessary investments to provide wholesale access will have to be undertaken by BBned. BBned will lease fibre from GNA and have the — […](27) — right to provide wholesale transport and related services to retail operators at its own risk. It will pay a fee per household connected to GNA for the use of the passive network.

(30)The wholesale operator will provide open, non-discriminatory access to retail operators to offer TV, telephony and broadband Internet access services. The services offered via the new network will compete with existing offers of cable and telecommunications companies such as KPN and UPC. Based on publicly available information, it appears that BBNed has already contracted with several retail service providers who have started to offer ‘triple play’ services(28) on the GNA fibre network to end users. The Commission notes that BBned is also active, through its affiliated companies Pilmo and Bbeyond, at retail level.

III. GROUNDS FOR INITIATING THE FORMAL INVESTIGATION PROCEDURE

(31)On 20 December 2006, the Commission decided to initiate the formal investigation procedure as the Dutch authorities did not provide sufficient evidence for the Commission to conclude on the conformity of the investment by the city of Amsterdam in GNA with the MEIP. The Commission considered it necessary to open the formal investigation procedure both for substantial and for procedural(29) reasons.

(32)In particular, the Commission was not convinced that the investment of the municipality of Amsterdam in GNA had been ‘concomitant’, i.e. it was performed at the same time with the private investments by ING RE and Reggefiber. Moreover, the Commission was not convinced that the terms and conditions of the investment by the municipality of Amsterdam in GNA were equal to the terms and conditions under which the other parties have invested. These concerns derived primarily from the fact that the municipality of Amsterdam spent money on the Citynet project before the establishment of GNA (so-called ‘pre-investments’)(30). Based on the preliminary assessment, the Commission had also doubts with regard to the feasibility of GNA’s business plan. Thus it was not possible at that stage to exclude the presence of State aid.

(33)Taking also into account these difficulties encountered, the impact of potential State aid on the investments of private operators, the opening of procedure was also justified because the Dutch authorities did not submit to the Commission all information necessary to assess the project.

(34)Finally, the Commission initiated the formal investigation procedure also in order to give the Dutch authorities and third parties the opportunity to submit their comments on its provisional assessment of the measure described and to make available to the Commission any relevant information related to the measure.

IV. COMMENTS SUBMITTED BY THIRD PARTIES

(35)UPC submitted its comments by letter registered on 17 July 2007. UPC’s submission reiterates its position expressed before the Commission decision to initiate proceedings(31) according to which the business plan of the Citynet project is based on unrealistic economic assumptions, in particular regarding penetration rates and wholesale prices, hence the business plan does not appear to be commercially viable. To substantiate this claim, UPC submitted a report prepared by a consultancy firm, RBB Economics, which updates other reports prepared by the same consultancy on the issue(32) submitted before the opening of the formal investigation(33).

(36)As regards the commercial viability of the project, UPC argues that the new fibre network would not enable operators to offer significantly different services from the services offered by the current operators. Consequently, and taking into account that the broadband penetration in the areas to be covered by GNA has already reached 65 % of the households, the new fibre network would fail to attract enough customers for its services to make its business plan viable.

(37)To support this argument, UPC submitted figures concerning the overall churn rates(34) for its own cable network concerning the areas where GNA is already providing(35) triple play services. The churn figures cover the periods from 1 January 2006 to 1 June 2007 in the areas where the GNA network is being deployed. The maximum churn figures indicate that UPC’s penetration rates for analogue cable TV, Internet access, telephony or digital TV in the concerned areas were stable, and were even increasing for Internet access. Therefore, UPC observes that there are no indications that there was a marked decrease of penetration due to the presence of GNA.

(38)In addition, UPC argues that the successful fibre deployments in the Dutch towns of Nuenen and Hillegom(36) brought forward by the Dutch authorities cannot be considered as a proper comparison with the Amsterdam project, as these projects allegedly received significant subsidies from the Dutch Ministry of Economic Affairs, the networks are jointly owned by the inhabitants and the operators, the projects are relatively small(37), and existing broadband operators only offered basic broadband services in both towns. This was in contrast with the situation in Amsterdam, where high speed offers are available on the existing cable and telecommunications networks.

(39)According to UPC, as GNA’s business plan is unlikely to be realistic, the investment of the Amsterdam municipality in GNA is not in line with the MEIP and amounts to State aid. By reference to the Commission’s decision in the Appingedam case(38), UPC also reiterates its prior position that any State aid involved in the project cannot be considered compatible with the Treaty as it would be neither necessary nor proportionate.

(40)Furthermore, UPC also expresses doubts as to whether the public investment in GNA is made on equal terms with the investments of the other shareholders of GNA. UPC argues that the fact that the municipality of Amsterdam undertook certain pre-investments and other GNA shareholders decided to commit themselves to participate in the project after the municipality performed certain feasibility studies indicates that the investment was not pursued on equal terms by all the shareholders which would shed further doubts on the market conformity of the measure.

(41)Finally, UPC also points out that if the project fails, the network might be sold cheaply to another operator, which will not have to bear the full investment costs. This operator therefore could strengthen its competitive position by capitalizing on a publicly funded project. UPC also highlights the alleged precedent character of the measure. In case the Commission were to find that the measure does not constitute State aid, other municipalities might follow the example of Amsterdam in deploying FttH networks. UPC highlights that the Dutch Ministry of Economic Affairs and an engineering firm, Arcadis(39) have already developed a freely available, on-line business model calculation(40) allegedly encouraging municipalities to develop their own FttH networks.

(42)ONO, a Spanish cable operator offering broadcasting transmission and telecommunications services, is of the opinion that the Commission should apply the MEIP test strictly as other municipalities might follow the example and invoke the MEIP in similar projects which could give rise to the proliferation of State subsidies. ONO asserts that public intervention should only take place in specific circumstances such as in the presence of market failure, ensuring the proportionality of the measure and respecting the principle of technological neutrality, which is allegedly not the case in the measure at hand. If the investment of Amsterdam constituted State aid, ONO argues that any aid involved would be incompatible with the common market, as the circumstances are similar to that of Appingedam(41), and in that case any aid granted will be subject to a recovery obligation. ONO points out the alleged precedent character of the decision on the question whether and under which conditions the rollout of FttH networks can be supported with public funding.

(43)France Telecom (‘FT’), the incumbent fixed line telecom operator in France, takes a similar view to UPC and ONO. The company expresses its doubts whether the investment of the municipality of Amsterdam in the Citynet project is in line with the MEIP. FT also supports the Commission’s approach in distinguishing State aid measures in ‘white’, ‘grey’ and ‘black’ areas, which is applied by the Commission when carrying out the compatibility assessment in the presence of aid(42). FT asserts that any State intervention in ‘black’ areas (such as Amsterdam) seems difficult to justify, as it results in serious distortions of competition. According to FT, this was particularly the case for the measure at hand, since the deployment of fibre access networks require heavy investments from market operators.

(44)The comments submitted by COM HEM, a Swedish cable operator, take a similar line to the other operators. COM HEM generally doubts the market conformity of the investment of the Amsterdam municipality. In its observations, COM HEM argues that public funding for broadband projects in urban areas are seldom in line with the MEIP and have strong distortive effects on competition. COM HEM also calls for a strict application of the MEIP and highlights the precedent value of the Amsterdam case to future municipal and other public sector investments.

(45)Another party requiring anonymity, which provides telecommunications services in several European countries, welcomed the Commission decision to initiate the formal investigation procedure. The company argues that its investment is endangered by the initiatives of municipalities using public funds which have a distortive effect on the market. The company expects the Commission to ban any public money in the Citynet Amsterdam project.

V. COMMENTS SUBMITTED BY THE DUTCH AUTHORITIES

(46)By letter registered on 16 March 2007, the Dutch authorities submitted their comments in connection with the Commission’s decision to initiate the formal investigation procedure. Furthermore, the Dutch authorities also provided comments on non-confidential versions of the studies prepared for UPC by RBB.

(47)Throughout their submissions, the Dutch authorities maintain their position that the investment of the municipality of Amsterdam in GNA is in line with the MEIP and therefore does not constitute State aid.

V.1. Comments submitted in connection with the decision to initiate procedure

General comments

(48)The Dutch authorities highlight that large-scale fibre deployments are taking place in the US and in Asia, and that Europe, in comparative terms, is lagging behind. The authorities claim that projects similar to the one under assessment are particularly beneficial for the European economy and fully in line with the Lisbon agenda(43).
(49)The authorities also emphasise the pro-competitive business model of GNA, which provides — contrary to the closed model of cable operators — open and non-discriminatory access to all retail operators. They argue that the new business model, inter alia, promotes service competition, boosts innovation and helps to reduce the risk of service providers by allowing them to use funding which matches the characteristics of each individual layer.
(50)The Dutch authorities take the view that the shareholders of the project provide further evidence that the GNA project is pursued on market terms. In this respect, the authorities emphasize the fact that two private investors and the commercial subsidiaries of the housing corporations are willing to participate in the project under the same terms and conditions as the municipality. They also refer to the open tender procedures for the construction of the network and for the wholesale service provision of the network. They further highlight that the fact that a significant bank loan was offered to GNA on market terms constitutes clear evidence that the project and the underlying business plan are based on prudent market terms.
(51)The Dutch authorities also stress that numerous similar successful fibre projects all over Europe and especially in the Netherlands(44) undertaken by market operators provide sufficient evidence that such investments can be pursued under normal market conditions.

On the pre-investments

(52)Regarding the pre-investments by the municipality of Amsterdam, the Commission noted in the opening decision that the municipality took initiatives before establishing GNA which seemed to go beyond what normal market practice would suggest. The Commission expressed concerns that the pre-investments might have reduced the risks associated with the project for all investing parties. Some of the start-up risk of the business underlying the GNA business plan might have been absorbed or mitigated by the municipality of Amsterdam before the investments by ING and Reggefiber in GNA were made. Based on the information submitted by the Dutch authorities before the opening of the formal investigation, it could not be clarified whether all shareholders in GNA did invest under the same terms and conditions.
(53)In their reply, the Dutch authorities underline that all investors committed themselves on 24 May 2006 to the investment in GNA on identical terms and conditions.
(54)As regards the pre-investments (reaching the amount of EUR […]), the Dutch authorities claim that it has always been the understanding of all GNA shareholders that the pre-investments by the municipality of Amsterdam would have to be repaid by GNA. To support this assertion, the Dutch authorities distinguish between two parts in the pre-investments.
(55)As regards the first part, the Dutch authorities stress that, although the agreements establishing the GNA were only signed on 24 May 2006 binding all parties to the EUR […] investment, in the letters of intent of […], the future GNA shareholders decided together to earmark a lump sum of EUR […] (ex. VAT) for the preparation costs of the project. Amongst others, this amount also covered the costs of the tender procedures for the selection of the builder of the network (BAM/Draka) and for the wholesale operator (BBNed), certain costs related to the notification procedure and certain digging activities.
(56)As for the second part of the pre-investments (amounting to EUR […]), the Dutch authorities argued that such pre-investments are fully in line with normal market practice: prudent market investors would follow the same practice in joint projects as one of the parties usually has to take on the role of ‘lead investor’. The authorities also stressed that pre-investments by the municipality of Amsterdam did not reduce the start-up risk of the investment for the other GNA shareholders.
(57)The Dutch authorities also argue that the pre-investments did not provide any advantage to any party, therefore could not constitute State aid within the meaning of Article 87(1) of the EC Treaty. For instance, the start of certain digging activities was triggered by the fact that some civil works took place in areas which were important for the future construction of the GNA network. Since all digging activities are coordinated in Amsterdam, digging just a few months afterwards would have been impossible, which would have caused delays and additional costs for GNA.
(58)Furthermore, the Dutch authorities claim that all these costs were incorporated in the business plan and that no new cost elements emerged that would not have been known by the other GNA shareholders.
(59)To support these claims, the municipality of Amsterdam submitted a report prepared by Deloitte — an accounting firm — of 18 January 2007, auditing the reimbursement of the pre-investments. The report states that the above-mentioned pre-investments, amounting to EUR […] were entirely reimbursed by GNA at the end of 2006 with interest would be charged. The Dutch authorities informed the Commission that the interest rate applied(45) is […] % in line with the Dutch national law(46), the interest amounts to EUR […] and were paid on […] by GNA.
(60)Furthermore, the Dutch authorities do not consider their own feasibility studies as part of the pre-investments. According to the Dutch authorities, all prudent market investors would carry out such studies. The Dutch authorities argue that these studies could not have reduced or absorbed some of the start-up risk for the other GNA shareholders, as the other potential shareholders have to follow their own appraisal as well and assess their own risks and benefits from the project.
(61)Similarly, the feasibility studies of the other GNA shareholders (such as ING RE’s or Reggefiber’s) carried out before investing into GNA were financed by the respective parties without being charged to GNA.
(62)Therefore, the Dutch authorities assert that the preliminary doubts of the Commission regarding ‘concomitance’ and ‘identical terms and conditions’ originating from the pre-investments of the Amsterdam municipality are properly addressed by the above-mentioned explanations provided and by the reimbursement of the relevant pre-financed costs to GNA.

Comments on the doubts related to the GNA business plan

(63)With regard to GNA’s business plan, the Dutch authorities argue that investments made by public authorities satisfy the conditions of the MEIP if they are pursued under the same terms and conditions as those made by private investors. The presence of private investors should guarantee that the project is done on market terms. Therefore, they argue that it was not strictly necessary for the Commission to analyse GNA’s business plan.
(64)Second, the Dutch authorities consider that all assumptions of GNA’s business plan were not optimistic, rather conservative ones.
(65)More particularly, regarding the financial indicators in the business plan, in the opening decision the Commission compared the targeted financial indicators with relevant publicly available comparative data, namely the weighted average cost of capital (‘WACC’)(47) of peer companies(48). The Commission came to the preliminary conclusion that overall, GNA’s targeted financial indicators do not seem unrealistic, but are highly dependent on the achievement of key success factors. Furthermore, the Commission claimed that the forecasts for the business plan’s key success factors are likely to be very uncertain due to the novelty of the Amsterdam project in terms of the technology used (fibre technology), the business model (three-layer model), the limited project size (hence only limited economies of scale) and the expected consumer demand for high bandwidth services.
(66)The Dutch authorities argue that the project’s internal rate of return(49) should not be compared with WACC figures of vertically integrated operators, as was done in the Commission’s opening decision as these figures include the risk of all three layers of these network operators. On the contrary, GNA is only investing in the passive network, which has an expected economic lifetime of 30 years (or even more) and which is more similar to infrastructure investments from which, generally, lower rates of return are required by investors.
(67)Regarding the targeted penetration rate, in the opening decision, the Commission came to the preliminary conclusion that, based on the available data, achieving at least […] % penetration for GNA’s ‘minimum scenario’ seems optimistic. Moreover, the target of […] % of all households after […] seems aggressive and will only be possible through a massive ‘penetration pricing’ strategy shifting existing customers of other operators to the GNA network.
(68)The Dutch authorities argue that there are already several highly successful fibre projects in the Netherlands, where fibre networks could achieve high penetration ratios. For instance, as regards Nuenen or Hillegom, the authorities claim that fibre penetration reached more than 80 % after one year of operation. Furthermore, the Dutch authorities argue that current broadband penetration rates are not relevant in the case at hand: services provided over fibre networks should be considered as a new market compared to current broadband offers of existing operators. Therefore, current broadband penetration rates should not be used as a benchmark. In addition, the authorities argue that numerous new services and applications(50) are emerging that require high speed symmetrical networks which can only be provided on a fibre-to-the-home network. Based on these arguments, the Dutch authorities stress that reaching approximately […] penetration rate in areas of Amsterdam where the network is rolled out within […] is not only realistic, but rather a conservative target.
(69)In relation to the wholesale prices charged by GNA to the wholesale operator, the Commission argues in the opening decision that although experts suggest that fibre networks entail lower operational expenditure than current copper telecommunications and cable networks, the GNA wholesale prices are still considerably lower than what market reports suggest.
(70)The Dutch authorities argue that fibre networks entail lower operational costs and have a longer economic lifetime than existing networks therefore lower wholesale prices are possible compared to what data on existing networks might suggest. This cost advantage provides sufficient room for relatively low wholesale prices compared to the offers on current copper networks.
(71)Regarding the validity of the investment cost figures, the Commission found in the opening decision that benchmark figures indicated that the capital expenditure per connection projected by GNA appeared to be low in comparison with data available from market players and other sources.
(72)The Dutch authorities argue that the feasibility of the planned investment costs are further underpinned by the topology and the characteristics of the geographic area where the network is deployed: the areas concerned in Amsterdam have a high population density and many newly built or renovated and multi-household buildings that help to reduce the costs of the deployment of the network per household.
(73)Regarding the appraisal of the residual value of the network, the Dutch authorities argue that the estimate in GNA’s business plan is realistic, as the economic lifetime of fibre networks can be 30 years or even more. Within this timeframe, no major additional investment or maintenance costs will be necessary, contrary to existing copper or cable networks. Furthermore, the authorities argue that the network in place will have significant ‘strategic value’ due to the ‘natural monopoly characteristics of fibre access networks’ that GNA will enjoy due to its first mover advantage.
(74)The Dutch authorities also commented on the Commission’s preliminary conclusion that all assumptions underlying the business plan seemed optimistic and that there is a high degree of sensitivity for the success of the project if even one of the targets (such as penetration grade) does not materialize.
(75)In this respect, the Dutch authorities had the feasibility of the assumptions reviewed and endorsed by a report by Stratix Consulting/Delft Technical University(51). The submitted report also assessed the feasibility of the planned GNA penetration rate. By analysing similar projects in and outside Europe and in particular in the Netherlands, estimating also the future market demand for high bandwidth symmetrical networks (i.e. for fibre networks), the submitted report argues that the planned GNA penetration figures can be considered as a conservative estimate. Furthermore, the report argues that following increased market demand for the services requiring fibre networks and also due to the pro-competitive nature of the open infrastructure, the demand for fibre networks may evolve even faster and to a higher level than predicted a few years ago.
(76)The main conclusion of the submitted study, shared by the Dutch authorities, is that the investment of the Amsterdam municipality is commercially viable and therefore in line with the MEIP.

V.2. Comments submitted in connection with Third Party observations

General comments

(77)First, the Dutch authorities reiterate their position that the doubts of the Commission regarding ‘concomitance’ and ‘identical terms and conditions’ originating from the pre-investments of the Amsterdam municipality have been properly addressed by the reimbursement of the relevant pre-financed costs by GNA. Furthermore, they claim that GNA’s business plan is feasible and realistic, therefore the investment of the Amsterdam municipality in the GNA project is fully in line with the MEIP.
(78)Second, the Dutch authorities also highlight again that, according to their view, in line with the existing case law(52) and Commission’s decisions(53), the significant participation of private investors in a project under identical terms and conditions as the public investor should be considered as ‘conclusive evidence’ that the MEIP is met and hence no State aid is involved.

Comments on Third Party observations

(79)The Dutch authorities, by referring to Council Regulation (EC) No 659/1999(54), argue that if an undertaking’s competitive position cannot be affected directly by the measure at hand, that party cannot be qualified as an interested party. Therefore, the Dutch authorities call upon the Commission not to take into account the comments of ONO, France Telecom(55) and Com Hem, since they do not have any business interest on the Dutch and particularly on the Amsterdam broadband market.
(80)The Dutch authorities argue that all parties providing comments in this case failed to take into account that the investments of private investors are sufficient to conclude the MEIP conformity of the investment. They also claim further that, according to the relevant Court jurisprudence, the analysis of the business plan was not necessary in the presence of these private investors.
(81)Regarding UPC’s claim that broadband penetration in the areas to be covered by GNA has already reached 65 %, therefore GNA would fail to reach the targeted penetration ratios, the Dutch authorities argue that the addressable market is much higher for GNA as not only broadband services, but TV, telephony and other new services will be offered to customers over GNA’s network. In relation to the churn rates submitted by UPC aiming to demonstrate the lack of sufficient interest for the services which can be provided over the GNA network, the Dutch authorities argue that the data is not relevant as service delivery over the network only started in March 2007, and therefore no conclusions could yet be drawn based on those data(56).
(82)For the reasons explained in recital 79, the Dutch authorities doubt the qualification of the four other companies as interested parties and are of the opinion that these parties only submitted general observations which are not relevant for the assessment of the measure at hand. Therefore, the Dutch authorities do not consider it necessary to comment these observations. The Dutch authorities also do not provide comments on the observations by third parties on the compatibility of any State aid contained in the measure as according to the authorities, a compatibility assessment is not necessary in the current case as no State aid is involved in the measure.

VI. STATE AID ASSESSMENT

(83)The Commission has examined whether the measure can be qualified as State aid within the meaning of Article 87(1) of the EC Treaty, which provides that ‘any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the common market’. It follows that in order for a measure to be qualified as State aid, the following cumulative conditions have to be met: 1) the measure has to be granted out of State resources and be imputable to the State, 2) it has to confer an economic advantage to undertakings, 3) the advantage has to be selective and distort or threaten to distort competition, 4) the measure has to affect intra-Community trade.

(84)When assessing the measure, the market context has to be taken into account: the electronic communications sector has been fully liberalized several years ago in Europe. In particular, investments in broadband networks in so-called ‘black areas’ where different broadband services are offered over at least two competing infrastructures (such as telephone and cable TV networks)(57) are primarily driven by private companies. The Commission’s policy in this field is to be stricter with regard to projects involving public funds in such areas because of the distortive effect on the business of private operators(58).

VI.1. State resources

(85)First, it has to be assessed whether the measure is granted directly or indirectly through State resources and imputable to the State(59).

(86)In the current case, the municipality of Amsterdam invests EUR 6 million in the partnership GNA. Since a municipality is to be considered as an emanation of the State, this is to be considered as an investment with State resources within the meaning of Article 87(1) of the EC Treaty(60). The Dutch authorities have not contested the existence of State resources.

VI.2. Advantage

VI.2.1.Market Economy Investor Principle

(87)For the purpose of establishing whether a financial investment by the State in an undertaking involves an advantage within the meaning of Article 87(1) of the EC Treaty, the Commission applies the MEIP in line with the relevant case law of the Court. Pursuant to that principle, a transaction does not involve State aid if it takes place at the same time and under the same terms and conditions that would be acceptable to a private investor operating under normal market economy conditions(61).
(88)According to established case law, capital placed by the state, directly or indirectly, at the disposal of an undertaking in circumstances which correspond to normal market conditions cannot be regarded as State aid. However, if the investment by a public investor disregards any prospect of profitability, even in the long term, such an injection must be regarded as State aid within the meaning of Article 87(1) of the EC Treaty(62).
(89)On this basis, a capital contribution from public funds must be regarded as satisfying the market economy investor test and not constituting State aid if, inter alia, it was made at the same time as a significant capital contribution on the part of a private investor made in comparable circumstances (the ‘concomitance’ test)(63).
(90)In the present case, the investment by the municipality of Amsterdam was made jointly with two private investors, namely ING RE and Reggefiber. The Commission will therefore first examine whether the investment of the municipality of Amsterdam fulfils the market economy investor test due to the fact that it was made concomitantly with a significant private investment. For this purpose, the Commission will examine the following criteria:
(91) First, it has to be identified whether these investors are market investors and whether the investments by the private investors have real economic significance. Such significance should be assessed in absolute terms (a significant portion of the total investment) and in relation to the financial strength of the private investor concerned.
(92) Second, it has to be assessed whether the investment by all parties concerned take place at the same time (‘concomitance’).
(93) Third, it has to be identified whether the terms and conditions of the investment are identical for all shareholders.
(94) Fourth, in cases where the State, other investors or the beneficiary have other relationships outside this investment (for example through a side-letter, providing for a guarantee by the State), there may exist grounds to doubt whether such equivalence in the mere investment terms suffices(64).
(95)Subsequently, and at a subsidiary level, the Commission will also examine GNA’s business plan, in particular in view of the claims of the complainant Liberty Global/UPC and other interested parties, which argued that there is no feasible business case for the network built by GNA.

VI.2.2.Assessment of the measure in light of the MEIP

VI.2.2.1.Significant participation by private investors
(96)As the Commission acknowledged in the opening decision(65), both ING RE and Reggefiber can be considered without any doubt as ‘private investors’(66). Additionally, both companies’ business interests are coherent with the project carried out by GNA: ING RE is active in the real estate business and infrastructure type of investments and Reggefiber is engaged in several fibre network projects in the Netherlands. The Commission also notes that the two different private operators can bring different types of expertise in order to contribute to the successfulness of the project.
(97)In absolute terms, the two private investors both invest a substantial amount (EUR 3 million, respectively) in the partnership GNA. If an investment of EUR 3 million, compared to the financial strength of both ING RE and Reggefiber, could be considered in relative terms to the size of the investors to be a small investment, such an investment is certainly significant in relative terms to the overall capitalisation of GNA and the capital contribution of the municipality of Amsterdam.
(98)Together the two private investors take up one third of the total equity. In the context of this project, this stake — while not giving the two companies outright control of GNA — is a significant portion of the overall investment. To this end, the Commission observes that both ING RE and Reggefiber are major shareholders of GNA. Indeed, the largest single shareholder (the Amsterdam municipality) owns 33 % of the equity of GNA. Each ING RE and Reggefiber have a stake equal to half of that of the largest single shareholder; if combined, they hold the same stake as the municipality. The remaining shareholders in the venture, the five housing corporations, hold singularly smaller stakes than ING RE and Reggefiber. It follows that within GNA there is no single shareholder capable of exerting majority control over the company. Moreover, the two private investors can be singularly, and even more if taken together, pivotal in forming a controlling majority within GNA(67).
(99)The Commission also notes that according to Dutch corporate law, one third of the shares are sufficient to form a blocking minority regarding any important decision of GNA. Therefore, the two private investors can jointly form a blocking minority within the company.
(100)To conclude, both ING RE and Reggefiber are market investors and their investments have real economic significance both in absolute and relative terms if seen in the context of the shareholding structure of GNA.
VI.2.2.2.Concomitance
(101)In the opening decision(68), the Commission acknowledged that the municipality invested de jure at the same time as the private investors in the partnership GNA, but expressed some preliminary doubts about the de facto concomitance of the investments of all GNA shareholders, as the municipality had already undertaken initiatives and investments before a final agreement with all other investors was concluded.
(102)The Commission noted that the municipality commissioned several studies in 2003 and 2004 to prepare the project. Furthermore, the municipality took initiatives which seem to go beyond these preliminary steps. Despite the absence of a firm commitment by private investors, the municipality published and organised tenders and even negotiated contracts for the construction and the exploitation of the network. In addition, the municipality of Amsterdam financed certain digging activities and purchased software for the construction of the network.
(103)In this respect, the Dutch authorities provided further information (audited by Deloitte) on the pre-investments and on their reimbursement of the pre-investments after the opening of the formal investigation. According to the Deloitte audit report, the commonly earmarked funding of the GNA shareholders amounted to EUR […] out of the total pre-investments of EUR […]. The remaining EUR […] were initially financed by the municipality only(69).
(104)The new information received during the formal investigation enabled the Commission to establish the following facts: First, a substantial part of the activities and pre-investments (EUR […]) was explicitly agreed in the ‘Letters of Intent’ signed by the prospective shareholders of GNA and co-financed by all of them proportionally to their stake even before the establishment of the company, as all partners in the venture considered necessary to undertake individually and separately certain steps prior to GNA’s incorporation.
(105)Second, all pre-investments (including EUR […] earmarked only by the Amsterdam municipality) were included in the business plan on which the investment was based and hence were agreed between all shareholders of GNA. Thus, all partners in the venture considered those pre-investments as useful steps prior to the establishment of GNA. According to the Dutch authorities, there was agreement among all shareholders that the portion initially financed by the municipality would be reimbursed by GNA. In other words, the municipality’s conduct did not pre-empt or influence the behaviour of the other market investors.
(106)On the basis of the above, the Commission considers that the fact that the municipality of Amsterdam did carry out some limited pre-investments prior to the formal setup of GNA does not call into question the fulfilment of the MEIP given that there was agreement among all shareholders that the municipality of Amsterdam would have to be reimbursed for these pre-investments.
VI.2.2.3.Terms and conditions
(107)In the opening decision(70), the Commission expressed preliminary doubts whether the terms and conditions of the investment for all shareholders were identical. More specifically, the Commission expressed its doubts that the pre-investments undertaken by the municipality of Amsterdam might have absorbed or mitigated some start-up risks associated with the project for other investing parties.
(108)These preliminary doubts derived from the fact that, although the Commission had requested several times a complete overview of the pre-investment costs(71), and the Dutch authorities submitted some information on the settlements between Amsterdam, GNA and its shareholders in this respect, this did not fully clarify all ‘pre-financing’ activities by the municipality including the amounts actually spent by the municipality of Amsterdam.
(109)Based on the submissions by the Dutch authorities at the moment the formal investigation was initiated, the Commission was not in a position to calculate or verify the total amount of these ‘pre-investments’ which the Commission estimated to be in the magnitude of EUR 1,5 million. Furthermore, it was not fully clear to the Commission how those costs were shared between the shareholders and included in the business plan of GNA.
(110)In particular, the Commission feared that the ‘pre-financing’ by the municipality — partly without the explicit agreement of some of the other investors — might have reduced the investment risk of the other investors in GNA and could have had a positive impact on their willingness to invest in GNA. The information submitted by the Dutch authorities following the opening of the formal investigation allayed these preliminary doubts.
(111)First, as clarified by the information provided by the Dutch authorities on the reimbursement of the pre-investment costs and by the submitted audit report, the total pre-investment costs amounted to EUR […], divided in two parts: EUR […] agreed and co-financed by all prospective shareholders of GNA in the ‘Letters of Intent’ proportionally to their stake, and EUR […] initially financed only by the Amsterdam municipality. All these costs were properly charged to GNA, including interest(72) and no additional GNA project costs seem to have remained unaddressed.
(112)Second, all costs related to the pre-investment were also initially included in the business plan of GNA, hence no new costs seem to have emerged after the signing of the GNA agreement that would have altered the terms and conditions for the other shareholders. This is in line with the finding that there was agreement among all shareholders that the municipality of Amsterdam would have to be reimbursed for the pre-investments carried out.
(113)Third, although the municipality of Amsterdam took the lead and pre-financed part of the project costs, those initiatives and investments could not have reduced the risks involved in the project for the other shareholders. Contrary to the preliminary doubts raised in the opening decision, the Commission comes to the conclusion that the investments initially financed by the municipality in undertaking limited civil works and purchasing software could not, in view of their nature and their limited financial significance, have altered the risk profile of the project. The same is true for the pre-investments co-financed by all prospective shareholders in parallel to the signing of the ‘Letters of Intent’. The analysis of the business plan clearly shows that the business risks of the investment in GNA is linked to the success of the project in the years to come, particularly in terms of market evolution, and not to the very first steps preparing the project.
(114)Fourth, the Commission does not agree with the argument of UPC that the feasibility studies conducted by the Amsterdam municipality might have reduced some risk for the other GNA shareholders, which might have decided to invest in GNA only afterwards. The Commission notes that any prudent market investor would normally carry out its own assessment on the strategy and profitability prospects (feasibility study) of an investment project(73) without this being considered a means to reduce the risks involved for other investing parties. In any case, if the project were to fail, all investing parties together would have to shoulder the losses resulting from an underperforming business or, in the worst case, the bankruptcy of GNA.
(115)In view of this information, the Commission concludes that all shareholders in GNA have invested under the same terms and conditions.
VI.2.2.4.Other relationships
(116)UPC argued that ING RE might have decided to invest in the GNA project not on the basis of economic considerations, but rather as part of their marketing strategy to maintain a good relationship with the municipality of Amsterdam, which is allegedly an important business partner of ING RE.
(117)It has to be assessed whether there are any other relationships outside the cooperation and investment agreement which are relevant for the assessment whether the investment meets the MEIP. In this respect, the opening decision(74) referred to the fact that the Dutch authorities have provided a statement in which they declare that there are no other relations between the parties concerned, i.e. relations outside the notified agreements, which are relevant for the assessment whether the investment is in line with the MEIP.
(118)The Commission’s own investigation and the information submitted by interested parties have not brought forward any elements indicating that the statement of the Dutch authorities is incorrect.
(119)On this basis, the Commission concludes that the investment by the municipality of Amsterdam was made at the same time and in comparable circumstances as the significant capital contributions by the private investors.
(120)Accordingly, the investment made by the municipality of Amsterdam is in line with the market investor principle and does not constitute State aid.
VI.2.2.5.Assessment of the business plan
(121)In addition, the Commission has also assessed the business plan of GNA, in particular in view of the doubts raised in its decision initiating the formal investigation and in view of the claims of the interested parties Liberty Global/UPC, ONO, FT, COM HEM and an anonymous party(75).
(122)In the opening decision(76), following the preliminary analysis of GNA’s business plan and the critical remarks received from UPC in this respect, the Commission considered that not only the targeted performance indicators, but all assumptions underlying the GNA business plan seemed optimistic. The Commission considered further that there was a high degree of sensitivity for the success of the project if the targets (such as the penetration rate) did not materialize, even more so if one of the underlying assumptions deviated from the targeted levels.
(123)The Commission received comments on the business plan analysis contained in the opening decision both from the Dutch authorities and from UPC which enabled it to deepen its assessment.
(124)The Commission notes that the business plan of GNA relates to a newly formed undertaking with no track record, acting in a new and innovative business segment (FttH technology, ‘three-layer model’, where passive and active infrastructures are operated and managed separately, with an open and non-discriminatory access offered to all retail operators, see description part recitals 21-30). Thus, in such a case, the assessment of a business plan of a newly formed undertaking is necessarily and inevitably based on future market projections and hypotheses regarding the likely evolution of demand and supply for FttH services.
Independent reviews conducted in relation to the business plan
(125)The Commission notes that the methodology of the business plan was audited and accepted by PricewaterhouseCoopers(77), an independent audit company. More specifically, after a first analysis of the first version of the GNA business plan, PricewaterhouseCoopers proposed some modifications to filter out certain preliminary inconsistencies. These proposals were taken into account in the updated GNA business plan. After those modifications, PricewaterhouseCoopers stated that the model does not ‘contain technical or economic integrity issues’, and the GNA business plan was submitted to the Commission afterwards.
(126)Following the Commission’s decision to open the formal investigation procedure, the Dutch authorities submitted a report prepared by Stratix Consulting/Delft Technical University(78), reviewing and endorsing the feasibility of the GNA business plan, with a special focus on the assumptions which were highlighted by the Commission as ‘optimistic’ in the opening decision. The report came to the conclusion that the GNA business plan is realistic and that the underlying assumptions are correctly based on current market trends and should be considered as ‘conservative’ estimates.
Financial indicators
(127)GNA applied three main financial performance indicators to measure the success of the project being the cash-flow generation, the return on equity and the internal rate of return (IRR).
(128)The Commission considers that the IRR ratio appears to be the most appropriate parameter for an analysis of the business plan. The IRR is used to make decisions on long-term investments and to compare different investment projects. The underlying IRR of GNA’s business plan is […] %.
(129)Given the novelty of the project and the dynamic nature of the broadband telecommunication markets, it is difficult to carry out a benchmarking exercise. The public availability of IRR figures from similar projects is limited, not least because these figures are rarely made public as they are considered business secrets. The most relevant publicly available comparative data is that on the weighted average cost of capital (‘WACC’) of other companies in the same industry. WACC data can be a useful benchmark because a project is considered worth undertaking if the IRR exceeds the WACC. Based on information in the possession of the Commission, industry figures fluctuate in the range between 8,1 % and 10,6 %(79). Based on these figures, the IRR in GNA’s business plan appears to be within the acceptable range.
(130)In its submission, UPC argued that due to the novelty and the high risk involved in the project, the targeted IRR for GNA, which is basically a start-up firm, must be higher than the WACC figures of a well-established company with significant customer base and cash flow generation, such as KPN or UPC.
(131)The Dutch authorities question the Commission’s assessment regarding the benchmarks used for the assessment of the IRR in the opening decision. They claim that the WACC figures are related to vertically integrated operators, while GNA is only investing into a passive network, which is more similar to an infrastructure investment for which lower rates of returns are generally required by investors.
(132)The Commission finds that the targeted IRR in GNA’s business plan appears to be within the market expectations for companies active in the telecommunications market. In addition, the Commission recognises that the investment project under scrutiny is different from that of a vertically integrated operator and it presents characteristics of an infrastructure type investment, for which a lower IRR is required.
(133)The Commission also assessed the alternative financial indicators used in the business plan, such as the positive cash flow generation and the return on equity. The Commission could not carry out a thorough benchmarking exercise for these indicators due to the lack of publicly available data(80), and therefore assessed them from the point of view of the adequacy and internal consistence within the business plan. Based on this analysis, the Commission concludes that the figures provided in the business plan appear to be acceptable to a market economy investor.
Penetration rate
(134)One of the most important factors for the business plan is the targeted penetration rate. As revenues for GNA will depend to a large extent on the achieved penetration rate, i.e. the percentage of connected households, the feasibility of achieving the targeted penetration rate is of crucial importance for the success of GNA’s business.
(135)GNA aims at achieving a penetration rate of […] % of all households within […] months (and penetration levels are expected to be saturated approximately at this level). In addition […](81).
(136)In its submissions, UPC expressed strong doubts whether any significant market share (such as the 40 % target rate mentioned in an investment report(82)) can be achieved by the project. UPC, relying on the report by RBB and figures concerning the overall churn rates for the areas where GNA is already providing triple play services, argues that the current overall broadband penetration in the areas planned to be covered by GNA has already reached 65 %. More particularly, according to UPC, the relatively strong competition in the area and the limited new offers available on GNA’s fibre network further supports the view that the retail operators using the GNA network via BBned will fail to attract enough customers.
(137)UPC also stresses that it does not see any competitive advantage of the new fibre network compared to the existing offers of current broadband operators other than the provision of symmetrical bandwidth to retail customers, for which service demand is however inexistent. Moreover, UPC argues that their most advanced broadband product in terms of available bandwidth(83) has not been very successful so far and that customers have a tendency of opting for cheaper and less-sophisticated products whenever more advanced products are offered in the market.
(138)In contrast, the Dutch authorities argue that new services can be provided over the symmetrical fibre broadband connection (such as file sharing and other ‘peer-to-peer applications’, HDTV), which belong to a different market than the current asymmetrical connections. Furthermore, the Dutch authorities relying on the already mentioned study prepared by Stratix Consulting and Delft Technical University stress that several successful examples not only from the Netherlands, but also from the US and Japan indicate that the targeted penetration rate is realistic.
(139)The Commission has assessed the arguments of the various parties in this respect. The Commission notes that GNA, in order to determine the penetration rate curve, […]. This penetration rate forecast model incorporates all the assumptions required, including take-up rates, speed of take-up rate and churn rates as well as those assumptions that were quoted by the UPC as being important factors to assess the feasibility of the business plan. Therefore, the business plan includes the relevant factors that are necessary for the assessment of the penetration rate.
(140)On the question whether the penetration rates in the business plan are achievable, the Commission notes that it will depend to a large extent on the evolution of the market, the speed at which new applications and technologies requiring very fast, symmetrical fibre broadband connection will be adopted by service providers and the reaction of customers to these new possibilities. These factors cannot be known with certainty at this point in time, as is demonstrated by the different views between UPC (more sceptical regarding market prospects) and Reggefiber, ING RE and Delft Technical University (who have endorsed the estimates in the business plan either as investors or as independent consultants).
(141)More specifically on the churn rates submitted by UPC to show that the project is failing to reach any meaningful penetration rates, the Commission notes that the submitted figures cover the period between 1 January 2006 and 1 June 2007, while retail operators using the GNA network started offering the services to customers in limited areas of Amsterdam only in March 2007. Given the very limited amount of time elapsed since services are provided over the GNA network, the Commission is not convinced that meaningful information could be elicited from the churn rates at this stage.
(142)In relation to UPC’s argument that the successful examples of small Dutch cities in fibre deployment, such as Hillegom or Nuenen(84) should not be taken into account as these cities are not comparable in terms of size, competitive situation and subsidies to the project of Amsterdam, the Commission considers that although there are certain differences, those projects still provide some useful insights for the benchmarking analysis as examples of fibre deployment projects in the Netherlands.
(143)Based on industry analysis(85) approximately 20-25 % of the population on average is willing to migrate to a new fibre network. […].
(144)Moreover, based on current trends, the demand for services being delivered over high bandwidth fibre networks seem to be increasing at a higher pace than what market experts expected a few years ago. The current plans for fibre networks deployment especially in the Netherlands and other European countries(86) provide further evidence for the attractiveness of investments in fibre access networks. Therefore, it is not unrealistic that a penetration rate of […] % of all households could be reached after […] months.
(145)The Commission therefore concludes that it is conceivable for a market economy investor to invest in the project on that basis of the penetration rates foreseen in the business plan (indeed, this is what Reggefiber and ING RE did).
Wholesale prices
(146)Wholesale prices per connection charged by GNA to BBned for the exploitation of the passive network […].
(147)In its submissions, UPC has calculated that GNA has to charge at least EUR 20-EUR 22 per month for a wholesale connection enabling triple-play services on its network. However, at this price level, the final retail service prices would necessarily be more expensive than current similar offers of the existing operators, which allegedly raises doubts about the targeted penetration rate.
(148)To support its arguments, UPC submitted calculations(87) for the wholesale prices based on the retail service prices offered by service providers(88) on GNA’s network. UPC argues that based on the current retail prices, it is unlikely that all three layers (passive, active and retail) will be able to operate profitably at the same time.
(149)In contrast, the Dutch authorities argue that the new fibre technology requires different cost calculations than the traditional copper or cable lines. For instance, the fibre network requires far lower operational expenditures (such as maintenance, management costs) and also has a longer economic life, i.e. a longer amortisation period (up to 30 years).
(150)UPC also asserted that it is not possible for all three layers(89) to be profitable simultaneously. In this respect, the Commission notes that it assessed the wholesale prices charged for access to the passive layer, as this is the layer included in the business plan under assessment, where public funds are involved. The Commission found that these wholesale prices are not unrealistic(90). The viability of the other two layers, operated by private operators, is to a certain extent affected by the wholesale prices charged for using the passive layer. Considering the fact that the operation of the second layer was won by BBned through an open tender, that an open, non-discriminatory access is offered to any operators on the retail layer and taking also into account the wholesale prices of GNA, the Commission does not see a risk that the two other layers might not be operated profitably.
(151)As regards the prices charged by GNA for the wholesale access, there are some indications that fibre networks may entail lower operational expenditure than legacy telecommunication and cable networks. Based on these considerations and also taking into account the benchmark information available and the results of the Stratix Consulting/Delft Technical University report, the Commission concludes that the wholesale prices in the business plan are not unrealistic compared to similar services provided by other operators.
Investment costs
(152)The Commission also assessed the planned investment costs. GNA calculates EUR […]. This leads to an overall cost for the passive layer of […] per household. The benchmark figures, collected by the Commission, for the complete infrastructure (active and passive together) fluctuate between EUR 1 000 and EUR 2 000(91), while the ratio between the investment needs for passive and active connection is roughly about two to one.
(153)UPC supports the preliminary view of the Commission expressed in the opening decision that the viability of the business plan is highly dependent on the investment costs and emphasizes that the calculated costs are very low estimates. In contrast, the Dutch authorities argue that the investment costs result from the topology of the areas concerned, the high population density and the presence of multi-household dwellings (lowering deployment costs per household). Moreover, the authorities highlight that GNA has already signed a contract with civil engineering companies Van den Berg/BAM and Draka Comteq in […] 2006(92) at specified conditions which should guarantee that the deployment costs will be in the range included in the business plan.
(154)Considering also the topography of the areas (densely populated area on a flat terrain), the fact that the civil engineering companies carrying out the works accepted to do them for a price compatible with the business plan, and the available benchmark data, the Commission finds that the planned GNA investment figures appear to be realistic. The Commission also takes note that GNA calculated with a certain level of buffer in case of an overrun of the investment costs which is in line with prudent market investors’ practice and provides further evidence for the robustness of the estimates of the investment costs.
Residual value of the network
(155)The estimated residual value of the passive network has a pivotal role in GNA’s business plan to achieve the financial targets and to provide a collateral for the investors. GNA applies a multiple-based calculation to determine the residual value of the asset (the projected value of the network in […] is EUR […]), the value is depending only on the achieved penetration rate. The Commission also assessed with an alternative multiple-based valuation the residual value of the GNA network, resulting in a value range of EUR […] million at […] % penetration rate, which matches GNA’s estimate.
(156)UPC suggests that the benchmark figures used by the Commission in the opening decision(93) for assessing the residual value of the network are too optimistic, since those figures are related to an established company with recurring revenues and customers. Furthermore, UPC asserts that for the expected economic lifetime of the network, an estimation of maximum 20 years appears reasonable (compared to the […] used in the business plan of GNA). UPC also recommends using a cash flow based assessment to check the residual value of the network.
(157)The Dutch authorities argue that the calculation of GNA’s business plan in relation to the residual value of the network is realistic and based on a correct methodology. In addition, they assert the economic lifetime of fibre networks can be […] or even more. Furthermore, they assert that the network in place will be valuable due to its first-mover advantage and because the construction of a second similar network might not be economically viable in view of certain natural-monopoly characteristics of fibre access networks. These latter factors allegedly provide further evidence that the network in place will have significant value even if the business plan does not fully materialize.
(158)Since limited relevant market data were available for this parameter, the Commission analysed the residual value of the network with several alternative methods, including a cash flow based assessment(94). Moreover, industry reports(95) support the view of the Dutch authorities that the economic lifetime of fibre networks may reach […]. Based on the available data, the residual value does not differ significantly from the price range derived from benchmark data. The Commission also takes note that the network is likely to be a tangible asset with significant value for the shareholders of GNA […](96).
Methodological concerns
(159)In the opening decision, the Commission also raised certain preliminary methodological concerns regarding GNA’s business plan. For instance, corporate tax is not included in the business plan. The Dutch authorities argue that the shareholders of GNA chose the legal form of ‘gesloten commanditaire vennotschap’ that has no obligation to pay profit tax. Therefore the results of GNA will be consolidated directly by its shareholders and the project is assessed on a ‘pre-tax base’.
(160)Having assessed the measure in more detail during the formal investigation, the Commission finds that the elements of methodological concern are of minor importance, based on the choice of the business plan methodology applied by GNA, and do not put into question the overall viability of the business plan. Furthermore, the Commission notes that both the private investors and also PricewaterhouseCoopers which audited the business plan endorsed(97) the methodological choices on which the business plan is based.
Other considerations regarding the business plan submitted by the complainants
(161)The Commission would like to point out that the Dutch authorities have also submitted several sensitivity analyses and business scenarios(98) to bolster the robustness of GNA’s business plan further. This information further demonstrated that GNA’s business plan takes into account all risks and opportunities related to the project.
(162)In its submissions, UPC criticised further assumptions of GNA’s business plan not explicitly mentioned in the assessment of the business plan outlined in recitals 121 and following, concluding that the plan would not be commercially feasible. The Commission would like to reiterate that neither UPC nor RBB, the consultancy which carried out several reports for UPC in the framework of the investigation of this case, had access to the actual business plan of GNA. Hence, their analyses were based on publicly available information, above all on an analyst report prepared on the feasibility of a fibre-to-the-home project in Amsterdam(99).
(163)The Commission notes that the main determinants of the business plan brought forward by UPC have either already been addressed in the Commission’s assessment of GNA’s business plan (such as penetration rate, subscriber take-up rate, churn rate, wholesale revenues, investment costs, life-expectancy of the network in the ‘residual value’ calculation) or are of limited relevance for its assessment (the choice of methodology in the plan)(100).
(164)UPC also stressed that the Commission should take into account not only the individual assumptions but the interrelation between them. For instance, UPC argues that high levels of penetration with high prices together are considered as impossible. The Commission has assessed the methodology and the figures on which the business plan is based and found that they can be considered acceptable to a private investor operating under normal market economy conditions.
(165)UPC also argues that in its calculations about the GNA business plan, it did not take into account that GNA will provide rebates for the users of the network in the first years, as this information was revealed in the Commission’s opening procedure. UPC argues that such contribution by GNA to the marketing costs of the complete project further deteriorate any financial return of the project for the investors and provide further evidence that the GNA business plan is not feasible. In this respect, the Commission notes that it is normal market practice that new services are offered at discount prices in the beginning in order to reach a certain customer base. Furthermore, the financial impact of those rebates was fully taken into account in the GNA business plan, which was analysed by the Commission.
(166)As regards the Arcadis model(101) indicated by UPC as a possible proxy for the GNA business plan, the Commission notes that the model was not part of the submission of the Dutch authorities, and the authorities confirmed that there is no link between the project under scrutiny and the Arcadis model.
(167)Furthermore, the Commission has no indication that the loan of EUR […] provided by […] was not provided on market terms. The interest rate required by […](102) and the potential collateral value of the asset in place provide sufficient evidence that the loan was granted on market terms.
Conclusion on the business plan
(168)After the formal investigation and having assessed the additional information received, the Commission comes to the conclusion that the GNA business plan is in line with the MEIP.
(169)The business plan is based on the main forecast that the evolution of the telecommunication and media markets in the Netherlands will translate in significant demand for high speed broadband services, which will be to a large extent be satisfied by FttH networks. From this forecast, which has been endorsed by the report of Stratix Consulting/Delft Technical University, derive the assumptions regarding prices and penetration rate of the GNA network. Together with other assumptions (e.g. investment costs), these factors drive the result that the project will sufficiently remunerate its investors.
(170)The Commission has assessed the plan and the assumptions therein by benchmarking it with comparable projects wherever possible, by checking its internal consistency and the feasibility of the assumptions. The conclusion from this assessment is that the business plan is methodologically correct and takes into account all the relevant factors that may impact on the success of the project.
(171)Furthermore, the Commission finds that GNA’s business plan does not contain manifestly erroneous assumptions. The assumptions underlying the plan are realistic if seen against the main forecast on the evolution of the telecommunications and media markets. They also appear realistic when benchmarked against comparable projects. The reservations that may be raised against some of the assumptions of the business plan as a matter of principle or starting from a different forecast about the future evolution of the relevant markets are not sufficient to undermine its overall credibility. Such divergence of views pertains to the wide margin of judgment that comes into entrepreneurial investment decisions.
(172)It should be underlined that two private investors specialised (also) in communications infrastructure investments, ING RE and Reggefiber, took a significant stake in the project on the basis of the business plan. […](103). In addition to the financial returns, the project’s strategic value (as a relatively small ‘pilot project’ to test the potential of FttH technology in densely populated areas and the underlying novel business model(104)) may well have contributed to the public and private investors’ decision to accept a relatively higher level of risk(105).
(173)Having these elements in mind, the Commission concludes that a private investor operating under normal market economy conditions could invest in GNA on the basis of its business plan.
(174)As regards the submissions of UPC, ONO, FT, COM HEM and the anonymous party, asking the Commission to adopt a strict line regarding the application of the MEIP in the present case in view of the alleged ‘precedent character’ of the project for other broadband investments in Europe, the Commission wishes to highlight that the MEIP assessment always has to be conducted on a case-by-case basis in line with the existing case-law and its decision-making practice in this field.
(175)Overall, the Commission therefore concludes that the investment by the municipality of Amsterdam does not constitute State aid since it was made at the same time as a significant capital contribution on the part of a private investor made in comparable circumstances. In addition, an examination of the business plan has equally confirmed that the investment is in accordance with the Market Economy Investor Principle.

VII. CONCLUSION

(176)Following the opening of the formal investigation procedure, the Dutch authorities addressed the preliminary doubts expressed by the Commission in the opening decision in a satisfactory manner. In particular, all pre-investments by the Amsterdam municipality, which were one of the main concerns of the Commission in the opening decision, were reimbursed by GNA. The Commission finds that the municipality of Amsterdam invests in GNA on the same terms and conditions as the private parties involved in the project, which are investing on the basis of GNA’s business plan.

(177)Consequently, the Commission concludes that the investment of the municipality of Amsterdam in GNA does not constitute State aid within the meaning of Article 87(1) of the EC Treaty as it is in conformity with the Market Economy Investor Principle,

HAS ADOPTED THIS DECISION:

Article 1U.K.

The investment by the municipality of Amsterdam in Glasvezelnet Amsterdam C.V., which is developing a fibre network in Amsterdam, as notified to the Commission by letter of 17 May 2005, does not constitute State aid within the meaning of Article 87(1) of the EC Treaty.

The measure may accordingly be implemented.

Article 2U.K.

This Decision is addressed to the Kingdom of the Netherlands.

Done at Brussels, 11 December 2007.

For the Commission

Neelie Kroes

Member of the Commission

(1)

OJ C 134, 16.6.2007, p. 9: State aid C 53/06 (ex N 262/05) — Citynet Amsterdam — investment by the city of Amsterdam in a fibre-to-the-home (FttH) network — Invitation to submit comments pursuant to Article 88(2) of the EC Treaty.

(2)

VECAI supplied further information by e-mails registered on 27 December 2005, 11 January 2006 and 31 January 2006.

(3)

UPC Nederland BV is a subsidiary of Liberty Global Inc. UPC Nederland BV acts under the name of UPC in the Netherlands. Liberty Global, Inc. (Liberty Global) is an international cable operator offering video, telephone, and Internet access services. It operates broadband communications networks in several countries.

(4)

Information supplied by UPC on 12 May and registered on 15 May 2006.

(5)

The authorities sent information on 19 and 31 May, 1, 7 and 13 June 2006.

(6)

The authorities sent information on 18, 25 and 29 August 2006.

(7)

Cf. based on publicly available information, e.g. articles in newspapers Parool (7 October 2006) and Trouw (13 October 2006).

(8)

The last batch of information was submitted by letter registered on 22 September 2006.

(9)

Minutes of the Court of Appeal dated 21 November 2006, No 200601252.

(10)

Gerechtshof, judgement in case 1252/06 KG of 18 January 2007.

(11)

‘Data rooms’ are typically used, for instance, during the due diligence phase of merger and acquisition transactions and provide access to confidential company data for the prospective bidders.

(12)

Cf. footnote 1.

(13)

By letter dated and registered on 17 July 2007.

(14)

By letter dated and registered on 16 July 2007.

(15)

By letter dated and registered on 17 July 2007.

(16)

By letter dated and registered on 17 July 2007.

(17)

By letter dated and registered on 17 July 2007, non-confidential version received on 27 July 2007.

(18)

The deployment of fibre access networks is considered to be the next big leap in the electronic communications sector. The new fibre networks when compared with existing copper based networks (such as ADSL or cable) will provide much higher speeds and symmetrical services and are expected to pave the way for numerous new, innovative services and applications based on IP technologies (IPTV, video on demand, telemedicine, etc.).

(19)

The successful bidder for the construction is Van den Berg Infrastructuren (BAM)/Draka Comteq Telecom. Van den Berg is a subsidiary of BAM, a major construction company and Draka Comteq is a producer of cables. The successful bidder for the operation of the network, tender 2005/S 79-76325, is BBned.

(20)

ING RE is a subsidiary of ING, a financial services (banking and insurance) conglomerate.

(21)

Reggefiber is engaged in several fibre network projects in the Netherlands and linked to the building and construction group Volker Wessels.

(22)

The financing of the social housing corporations in the Netherlands is currently subject of an existing aid procedure carried out by the Commission (Case E-2/2005 Bestaande woonwet en financierings-methoden voor woningsbouwcorporaties). By letter dated 20 December 2005 to Commissioner Neelie Kroes, Vecai requested the Commission to investigate the legal status of the housing corporations under the State aid rules. By letter of 3 February 2006, Commissioner Kroes replied, explaining that Competition DG is dealing with the issue within the framework of the before-mentioned ongoing procedure.

(23)

About one third of the houses (13 000) in this area are owned by the social housing corporations.

(24)

[…] The information in brackets is covered by the obligation of professional secrecy.

(25)

[…]

(26)

BBNed is a private broadband operator, a subsidiary of Telecom Italia. BBned was selected through a tender procedure which was advertised in the Official Journal of the European Communities with reference to 2004/S 138-118456, dated 17 July 2004.

(27)

[…]

(28)

In telecommunications, ‘triple play’ is a term for offering high speed Internet access, television and telephone services over a single broadband connection.

(29)

The Judgement of the Court of First Instance in case T-73/98 Société chimique Prayon-Rupel SA v Commission [2001] ECR II-867, recital 93 says ‘the fact that the time spent considerably exceeds the time usually required for preliminary examinations under Article 93(3) of the Treaty may, with other factors, justify the conclusion that the Commission encountered serious difficulties off assessment necessitating initiation of the procedure under Article 93(2) of the Treaty’.

(30)

The municipality commissioned several studies in 2003 and 2004 to prepare the project and organised the tenders for the construction and the exploitation of the network. In addition, the city of Amsterdam financed certain digging activities and purchased software for the construction of the network. These ‘pre-investments’ of the municipality amount to […].

(31)

See recital 30 and following of the opening decision.

(32)

As the Dutch authorities requested that GNA’s business plan was subject to the obligation of professional secrecy, RBB Economics did not have access to GNA’s actual business plan and based its study on publicly available information and on the public version of the Commission’s opening decision.

(33)

See recitals 30 and following of the opening decision.

(34)

The overall churn rates indicated all customers who left UPC to any other operator (for instance to KPN, GNA, moved to other districts, etc.). Therefore, UPC considers this figure as the maximum churn rate that the new GNA fibre network could have achieved at this stage.

(35)

GNA started to roll out the network in October 2006 and retail operators using the GNA network via BBNed started offering services in March 2007.

(36)

For instance, in certain parts of Nuenen and Hillegom, market penetration of services over the fibre network is higher than 80 % of all households after one year of service provision. The projects in Nuenen and Hillegom have a different setting than the Amsterdam project, for instance no public investor is present in Hillegom.

(37)

The Nuenen fibre network passes 7 400 homes, less then 25 % of the current size of the Amsterdam project.

(38)

See Commission Decision of 19 July 2006, C 35/05, Broadband development Appingedam (OJ L 86, 27.3.2007, p. 1), where the Commission considered aid for the development of a optic fibre access network to be incompatible with the common market as the Dutch municipality of Appingedam developed its network with State aid where private broadband network operators in the area already offered similar services. In the Appingedam case, there was no private involvement in the financing of the project and the MEIP was not invoked by the Dutch authorities.

(39)

Arcadis is an engineering company that provides project management, consultancy and engineering services for fibre network deployments.

(40)

The model is available at the following website: http://ngn.arcadis.nl/

(41)

See footnote 37.

(42)

The distinction between ‘white’, ‘grey’ and ‘black’ depends on the level of existing offers of broadband services. In general terms, ‘white areas’ are rural and scarcely populated zones with no broadband provision at all; ‘grey areas’ are zones where basic broadband services are already provided in some parts of the concerned territory; and ‘black areas’ are zones where different broadband services are offered over at least 2 competing infrastructures (such as telephone and cable TV networks).

(43)

Communication from the Commission to the Council and the European Parliament. Common Actions for Growth and Employment: The Community Lisbon Programme, COM(2005) 330 of 20 July 2005.

(44)

In the Stratix Consulting/Delft Technical University report prepared for the Dutch authorities on 8 March 2007 and submitted to the Commission on 16 March 2007, the consultant company lists approximately 50 ongoing fibre deployment projects in the Netherlands as of December 2006, with a similar amount of new projects announced and planned to be launched from 2007.

(45)

Information submitted on 9 November 2007 and on 12 November 2007.

(46)

[…]

(47)

The weighted average cost of capital (WACC) expresses the relative cost of equity and debt capital of a business. WACC is a widely accepted financial indicator to measure for any particular business or project the rate of return required by the providers of capital (both debt and equity) having regard to the risk characteristics inherent in the project. Businesses or projects which are able to earn returns (measured e.g. using the internal rate of return) greater than the cost of capital add value for investors. Conversely, businesses or projects which, while they may still be profitable, produce returns less than the cost of capital ‘destroy’ investor value.

(48)

In the Commission’s benchmarking analysis, the assumptions of GNA’s business plan were compared with the available information of its considered peer companies, mainly European telecom operators.

(49)

The internal rate of return is used to make decisions on long-term investments and compare different investment projects.

(50)

Such as peer-to-peer applications, file downloading, HDTV services, etc.

(51)

See footnote 44.

(52)

The Dutch authorities are referring in particular to the CFI judgement of 12 December 2000, Alitalia v Commission, T-296/97, recitals 80-81, Court Judgement of 21 March 1991, Italy v Commission, C-303/88, ECR I-1433, recital 20; and CFI judgement of 12 December 1996, Air France v Commission, T-358/94, ECR II-2109, recital 70.

(53)

The Dutch authorities are referring to Commission Decisions 95/404/EC on a procedure relating to the application of Council Regulation (EEC) No 2407/92 (‘Swissair/Sabena’) of 19 July 1995, and N 172/00 Seed and Venture Capital Scheme, Ireland of 17 October 2000 (OJ C 37, 3.2.2001, p. 48).

(54)

Council Regulation (EC) No 659/1999 of 22 March 1999 laying down detailed rules for the application of Article 93 of the EC Treaty (OJ L 83, 27.3.1999, p. 1).

(55)

France Telecom has recently sold its subsidiaries, Orange Netherlands and Nederland Breedband to Deutsche Telekom. The transaction was approved by the Commission on 20 August 2007.

(56)

The last data for the churn figures submitted by UPC relate to the situation as of 1 June 2007.

(57)

See also footnote 41.

(58)

See for instance Commission Decision of 19 July 2006, C 35/05, Broadband development Appingedam (OJ L 86, 27.3.2007, p. 1), also footnote 37.

(59)

C-345/02, Pearle and others [2004] I-7139, recital 35 with reference to Case C-303/88 Italy v Commission [1991] ECR I-1433, recital 11 and C-482/99, France v. Commission, Stardust Marine [2002] ECR I-4379, recital 24.

(60)

Legally, the municipality does not invest itself in GNA but through a special purpose vehicle that has been set up for this project […] and which is owned by the municipality of Amsterdam via the Development Corporation Amsterdam (OGA, an undertaking owned by the municipality of Amsterdam). The investment therefore stems from State resources which are imputable to the State (the subsidiary is owned by the municipality and the decision to invest itself has been taken by the municipality and is channelled through its subsidiary upon initiative by the municipality). Therefore, the conditions outlined in case C-482/99, recital 37 Stardust Marine (see footnote 58) have been met.

(61)

See for instance Judgment of the Court of 8 May 2003 In Joined Cases C-328/99 and C-399/00: Italian Republic and SIM 2 Multimedia SpA v Commission [2003] ECR I-4035, recitals 37-38, ‘Seleco judgement’; Joined Cases 296 and 318/82, Netherlands and Leeuwarder Papierwarenfabriek BV v Commission [1985] ECR 809, recital 17; Application of Articles 92 and 93 of the EC Treaty and Article 61 of the EEA Agreement to State aid in the aviation sector, op. cit. recitals 25 and 26; Commission Decision of 2 August 2004 (2006/621/EC) on the State Aid implemented by France for France Télécom (OJ L 257, 20.9.2006, p. 11); Communication of the Commission to the Member States 93/C 307/03 on the application of Articles 92 and 93 of the EEC Treaty and of Article 5 of Commission Directive 80/723/EEC to public undertakings in the manufacturing sector (OJ C 307, 13.11.1993, p. 3), recital 2.

(62)

See for instance Judgment of the Court in case C-303/88 Italy v Commission [1991] ECR I-1433, recital 20, ‘ENI-Lanerossi judgement’; Case T-358/94 Air France v Commission [1996] ECR II-2109, recital 70.

(63)

Case T-296/97, Alitalia [2000] ECR II-3871, recital 81; case T-385/94, Air France, [1996] ECR II-2109, recitals 148-149. Cf. equally the Commission’s Position, The Application of Articles 92 and 93 of the EEC Treaty to public authorities’ holdings, Bulletin EC-9/1984, para. 3.2(i) and (iii).

(64)

In other words, the terms and conditions can be identical in one agreement but, at the same time, other agreements can lay down additional clauses with different rights and obligations.

(65)

Footnote 48 of the opening decision.

(66)

ING’s shares are 100 % floated on the stock exchange without any shareholder owning more then 5 % of its shares. Reggefiber is a subsidiary of Reggeborgh which is an investment vehicle of the family Wessels. See also footnotes 20 and 21.

(67)

The significance of private participation also has to be seen in the context that broadband investment in black areas are primarily driven by private operators. See also recital 84.

(68)

Paragraphs 49 ff. of the opening decision.

(69)

See also recitals 54-55.

(70)

Recitals 52 ff. of the opening decision.

(71)

For instance in its letter to the Dutch authorities of 29 September 2006.

(72)

See also recital 59. The Dutch authorities informed the Commission that the interest charge amounted to […] and was paid on […] by GNA.

(73)

See for instance Commission Decision of 7 June 2005 on Alitalia’s restructuring plan (OJ L 69, 8.3.2006, p. 1), recital 194: ‘In order to present its offer, Deutsche Bank carried out an assessment of the company’s strategy and profitability prospects. In addition, before concluding the final contract, it proposes to carry out a due diligence operation which any investor should carry out before initiating the operation’.

(74)

Recital 62 of the opening decision.

(75)

In the following assessment, reference will be made mostly to UPC, as their submissions are the most detailed. However, the assessment also covers the more general remarks submitted by other interested parties.

(76)

Recitals 63 ff. of the opening decision.

(77)

Performed on 2 May 2006 on behalf of the Dutch authorities, submitted to the Commission on 11 May 2006.

(78)

Submitted on 16 March 2007, see also recital 75.

(79)

UPC (Dutch cable operator) 10,6 %, Fastweb (Italian broadband operator) 9 %, Telenet (Flemish cable operator) 8,5 %, KPN (Dutch incumbent) 8,1 %.

(80)

It should be noted that in its submissions UPC also did not use these indicators for its observations on GNA’s business plan.

(81)

[…]

(82)

UPC/RBB conducted their analyses on publicly available information, above all on an ING analyst report prepared on the feasibility of a fibre-to-the-home project in Amsterdam: ING analyst report ‘European Telecoms’, 24 February 2006.

(83)

Namely Chello Extreme, currently offering 20 Mbps download and 2 Mbps upload speed.

(84)

Where fibre penetration has reached more than 80 %.

(85)

See for instance JPMorgan (2006) study: The fibre battle — Changing Dynamics in European Wirelines, 4 October 2006. JPMorgan also calculates that the payback time of a fibre to the home infrastructure — providing open access in an average metropolitan market — starts looking attractive from a market share of 25 %.

(86)

As for instance evidenced by the investments of Reggefiber, Casanet in the Netherlands, Iliad, Free or France Telecom in France and examples from other European countries. See also: The Netherlands: FTTH deployment overview, 4Q2006 prepared by Stratix, January 2007.

(87)

Source: Supplement to the RBB report on Amsterdam’s investment in the FttH Citynet project, 7 November 2006.

(88)

Undertakings such as Pilmo, Unet or InterNLnet are already offering services on the GNA network with a limited geographical scope.

(89)

For a description of the three layer model see recital 23 ff.

(90)

For the assessment of the wholesale prices see from recital 146 ff.

(91)

For the whole infrastructure (active and passive) UPC/RBB estimates EUR 1 500 per home passed (EUR 1 000/EUR 500 respectively), KPN EUR 1 300, Hillegom (Dutch FttH project) EUR 1 200, Corning (fibre optic manufacturer) EUR 1 200, Arthur D. Little (consulting company) EUR 1 000 (EUR 600/EUR 400 respectively), Fastweb (Italian broadband operator) EUR 1 200, ARCEP (French regulator) EUR 2 000 (data from 2005 and 2006), JPMorgan (consulting company) EUR 1 000-EUR 2 000.

(92)

See also footnote 19.

(93)

Footnote 37 of the opening decision.

(94)

The cash flow based assessment (in line with the comments of UPC) resulted in a value range of EUR […].

(95)

See for instance FttH Council documents available at http://www.ftthcouncil.org/ or Gartner research of 14 February 2006: Governments Can Bring Moore’s Law to Broadband Access.

(96)

The Dutch authorities have also submitted several sensitivity analyses and different business scenarios as part of the GNA business plan. The Dutch authorities argue that even in the ‘worst-case scenario’ the value of the asset will be sufficient […].

(97)

Performed on 2 May 2006 on behalf of GNA, submitted to the Commission on 11 May 2006. See also recital 125.

(98)

In case of a business plan, sensitivity analysis enables for decision makers to assess the dependency of the financial outcomes of the business plan to the changes in the underlying assumptions. Similarly, different business scenarios also highlight different outcomes of the project in light of different market assumptions.

(99)

The ING report came to the conclusion that a fibre-to-the-home network in Amsterdam is financially feasible and viable (ING analyst report ‘European Telecoms’, 24 February 2006). However, the business plan used in the analyst report is significantly different compared to that of GNA (e.g. methodology, projected timeframe, financial targets, etc.).

(100)

GNA applies a different business plan methodology compared to the one applied in the ING report mentioned in footnote 98. In GNA’s methodology, a cost of capital indicator was not required. However, in its analysis, the Commission also used the cost of capital figures to benchmark the IRR rate of GNA’s business plan. See also recital 134 and following.

(101)

See recital 41.

(102)

The authorities provided the Commission with loan agreements on […] and it is stated that the interest charged is equal to […] %.

(103)

[…]

(104)

Namely the ‘three-layer model’, where the passive and the active layers are operated and managed separately, with an open and non-discriminatory access offered to all retail operators, see also Figure 1.

(105)

Small risky investments can also be explained by real option theory, according to which risky investments can entail a two-stage process. In the first stage, only a small investment is made (i.e. the company is purchasing an option). Some time later, when more information is known, the company can make a larger investment (i.e. exercising this option) or abandon its plans, thereby limiting its losses to the initial small-scale investment (i.e. to the costs of the option). By shouldering higher risk and uncertainty in the first stage (with small commitment), companies can reduce the risks of subsequent (larger) investments.