Public Consultation on a Review of the functioning of Regulation (EC) N 544/2009 (the "Roaming Regulation")

Transcription

1 Public Consultation on a Review of the functioning of Regulation (EC) N 544/2009 (the "Roaming Regulation") Orange France Telecom Group answer Public version February 11 th 2011

2

3 Executive summary: Orange France Telecom Group (OFTG) welcomes the opportunity given by the European Commission (EC) to express its views on the functioning and the possible extension of the Roaming Regulation 1. Roaming: an advanced and developing customer service Our customers find the current roaming service easy to use and it provides them with seamless coverage abroad. They are able, throughout the EU and worldwide, to benefit from many complementary services such as unified messaging systems; a single bill; the auto-completion of country codes when calling home; simplified access to service and a fully personalised and secured data environment. In addition, customers also benefit from growing data-use substitute technologies such as Wifi. And the service is constantly developing. OFTG has constantly decreased its roaming tariffs. OFTG has, for example, introduced over the last three years Orange Travel with a range of offers such as favourite countries (where calls are charged at discounted or domestic rates in the chosen countries) and data daily, with discounted data access for one day. Specific features of data Mobile data use has grown nearly 100 per cent yearly, requiring fast network deployment to reduce network congestion. OFTG is proud to have helped popularise data usage with attractive national bundle offers. For the last two years in Western Europe the penetration rate of smartphones increased from 5% to 50%, thanks notably to smarthphone subsidies. OFTG is extending this trend to its European roaming service to support the success of smartphones, tablets and dongles. In 2010, one million OFTG customers used Orange travel data daily offer (e.g. 2 / 2MB) across all Europe. Data tariffs are evolving from unlimited data to segmented bundles, showing the relative immaturity of the market in terms of business models. Consumer protection OFTG has fully embraced the aim of consumer protection. After having anticipated some of the regulatory measures, it has fully implemented them, sometimes even increasing tariff transparency beyond the requirements. OFTG has decreased customer roaming bill-shocks by several factors between 2009 and We will continue to improve customer communications, and provide specific information on the price of roaming. Moreover, in order to better protect occasional roamers from bill-shock, OFTG is activating a default data-roaming package for all new residential customers taking a smartphone. Roaming regulation and the European market place As suggested by the text of the consultation, OFTG has analysed numerous alternatives to roaming, including those considered by the EC. All were studied closely, focusing on their impact on the customer experience, regulatory efficiency and implementation. After careful review, OFTG has not been able to identify any alternative which would match current roaming arrangements in terms of technical efficiency and customer friendliness. Our conclusion so far is shared by BEREC, which, in its report 2, said: "none of the alternative measures has sufficient advantages to outweigh its disadvantages". 1 Roaming Regulation (EC) N 544/ BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58

4 While OFTG understands the EC willingness to enhance the Internal Market, OFTG doesn t believe in fully harmonised roaming prices for the following reasons: Market segmentation: domestic mobile markets have been and remain competitive; one reason for this success is the segmentation of the offers. This meets customer needs by employing a broad range of price tags for the different components of an offer, allowing him or her to choose the services that matter to them. Variation in provision costs: as acknowledged by BEREC, domestic markets naturally present distinct features (geography, cost of spectrum, coverage or quality of services requirements, market preferences) which directly influence the level of prices paid by consumers. These explain, in great part, the price differences amongst Member States. In 2011 and 2012, we can expect these national differences to be exacerbated with the spectrum auctions announced in a large number of member States. Competition in those national markets will not and cannot itself deliver, in the short run, equal prices across the European Union. Impact on national competition Any regulatory roaming approach should be carefully assessed for its effects on national competition. In particular, OFTG fully supports BEREC s concern on the possible damage caused by a forced convergence of roaming and domestic tariffs. Any regulatory measures - whose effects would be to impact and distort domestic markets - would go far beyond the objective of roaming regulation, raising questions of proportionality. Similarly, roaming regulation should not result in mobile access regulation, which the Commission itself considered disproportionnate in domestic markets when it adopted its last Recommendation on relevant markets back in A thorough impact assessment should take account of both the roaming market and domestic markets, before determining any regulatory remedies. This is also true if regulation is once again based on price-caps. OFTG will gladly participate in any study the Commission deems necessary. Proportionate glide path with a review in 2014 If regulation is to be maintained between 2012 and 2014, the Commission should consider a continuation of the glide path with proportionate decreases. The regulation based on price-caps must remain proportional and allow sufficient space for innovation and retail offer segmentation. OFTG has some important remarks regarding the methodologies and cost results presented in BEREC report on roaming. Any roaming regulation beyond mid-2012 should treat voice, SMS and data differently; the retail data market is still far from mature and the competition conditions in play will be further enhanced by Wifi or new technologies such as LTE. OFTG expects the data market growth to be supported by an increase of competitive roaming offers in the interest of the different customer profiles by 2014, as has been happening recently. OFTG therefore fully supports BEREC s proposal of a review in 2014, allowing the EC to take this evolution into account when considering the roaming key performance target of the Digital Agenda. 4

5 I. Markets development since the commencement of Regulation Question 1: To what extent do you believe that the current regulation achieved its objectives in terms of: (a) Contributing to the single market for roaming services? (b) Ensuring consumer protection? (c) Promoting competition? Please explain and substantiate your responses with data where possible. When looking at the roaming markets, it is obvious that over the past years, they have substantially evolved in terms of consumer protection and competition, contributing to harmonisation at European level and therefore to the single market. Strong price decreases over the past few years From a price perspective, major price decreases have occurred in the roaming markets since 2007, as can be seen in BEREC benchmarks 3. While tariff regulation has played a substantial role in such evolution, it is not the only reason justifying those decreases. Wholesale market are particularly competitive: contrary to some statements, OFTG observes strong competition when negotiating deals, whatever the size of the competitor, based on coverage, steering capabilities and prices. For voice wholesale, BEREC benchmarks indicate that prices went from 0.46 /min down to /min, demonstrating a 54% decrease of wholesale prices since Q OFTG average wholesale prices for data is far below the roaming regulation caps [CONFIDENTIAL] and unbalanced voice and SMS can also be far below, [CONFIDENTIAL]. In the case of retail voice, average prices for outgoing roaming calls have decreased from 0.69 /min in Q2 2007, when the first roaming regulation was introduced, down to /min in Q In the same period, the retail price of received roaming calls has decreased from 0.34 /min down to /min. In other words, over a three-year period, BEREC figures show that the average retail price of roaming voice calls in Europe has decreased ed by 45% for calls made and by 52% for calls received. Over the same period, Eurostat figures indicate that inflation was approaching 7% in Europe 4. 3 Unless otherwise indicated, all price levels in the following four paragraphs are to be found in the International Roaming ERG Benchmark Data Report for April September 2007 published in January 2008 for figures relative to Q2 2007, and in the International Roaming BEREC Benchmark Data Report for January 2010 June 2010 published in October 2010 for figures relative to Q More precisely, the figures can be found at the following pages of these respective reports: Voice Voice Retail Voice Retail SMS Data Wholesale outgoing incoming Retail Retail Q p14, Fig.7 p8, Fig.1 p9, Fig.2 p12, Fig.5 p13, Fig.6 Q p4, 1.10 p4, 1.6 p4, 1.6 p4, 1.13 (*) (*) Since the average retail data price presented by ERG for Q was apparently averaged over onnet and off-net traffic, consistency required to take the corresponding figure for Q This figure can be found in Table 1 on p6 of the BEREC report on International Mobile Roaming Regulation published in December Eurostat Harmonised Indices of Consumer Prices dataset, reference TEC As described by Eurostat, Harmonised Indices of Consumer Prices (HICPs) are designed for international comparisons of consumer price inflation. HICP annual average indices for EU evolved from in 2007 to in

6 Regarding SMS wholesale and retail, prices are set competitively and are not particularly higher or even at the same level as domestic SMS prices. The same BEREC benchmarks as above show that retail SMS prices decreased from 0.29 /SMS in Q down to /SMS in Q2 2010, thus indicating an average 67% decrease of retail SMS prices over a three-year period. In addition, and as at home, customers do not pay for receiving SMS whilst roaming. On data roaming, as acknowledged by BEREC, significant price decreases have occurred: The picture is different for data roaming, in that the average wholesale and retail charges have fallen significantly between the second quarter of 2007 and the second quarter of This is backed by BEREC benchmark figures, which showed an average retail data price in Europe of 5.81 /MB in Q dropping down to /MB in Q , representing nearly a 78% drop in the retail price of roaming data in three years! OFTG underlines that this substantial drop was achieved without any price regulation of retail data roaming. The sharp decreases have had a significant impact on the sector. Based on figures obtained by GSMA from AT Kearney, the mobile operators roaming revenue decreased 42% following the introduction of the roaming regulation, accounting for ~3.7 G of revenue loss over the whole European mobile industry. These figures have also to be read in the light of the global context characterized by investments in new spectrum band and networks, the current economic crisis impacting notably the business segment, and other regulation such as MTR decreases. New innovative retail offers Roaming is an advanced and developing consumer service, both in terms of services granted and design of retail offers. Roaming services grant currently end users easy and enriched services since in addition to making/receiving calls or SMS as well as surf on the Internet as easily as at home, end users also benefit of voice call, single bill, auto completion of country codes when calling home, secured data environment, etc. To develop notably those services, OFTG has invested in research and has filed dozens of patents on roaming-related ideas. To answer consumer s expectations and needs, retail mobile markets are also characterised by strong segmentation of retail offers in order to answer specific customer needs and expectations. Such features are essential in terms of competition and customer satisfaction. In particular roamers do not travel throughout the EU 27: most roamers have only a few preferred countries, not necessarily within EU 27, in which they roam during the year. Customers are thus interested, and best served, by aggressive offers on their preferred routes. A one size fits all approach would thus not allow an optimal matching with customer needs; operators need flexibility to innovate and offer retail offers answering the different type of consumers. In that sense, operators have developed new and innovative offers in terms of roaming. For example within OFTG: o o o Travel Data Daily: daily data offer when roaming launched by OFTG member companies (MCO) leading for example: Mobistar: 2 for 2MB with a price discount of 50% for any additional MB during pass validity; Orange France: Travel Data Daily is automatically activated for smartphone subscription, meaning once a customer with a smartphone travels within Europe and starts using Internet, then he will be granted 10MB for 5 per day; 5 International Roaming BEREC Benchmark Data Report for January 2010 June 2010, p 5, BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, Table 1, p6, Average retail price for data, including prepaid and postpaid, on-net and off-net 6

7 o o o o Orange France: International Internet passes valid during 7 days: 2 for 2 MB, 5 for 10 MB or 15 for 50 MB. Orange France s Travel pass valid 7 days: 5 inc. VAT: 15min (incoming and outgoing calls) + 10 SMS + 2MB within Europe USA Canada. See Appendix I for further detail Favourite country: it provides discounted or national rates for roaming calls and discounted, or free bundles of incoming calls, for an upfront monthly fee. For example, in France, this offer is granted at a monthly fee of 5 per month that results in national rates for making calls and in one hour of free incoming calls. Such an option can be extended to all EU countries with a 10 /month fee. In Belgium, this option is sold 3 per month and results in a reduction of 33% of calls made and 17% of call received, such reductions being increased respectively to 57% and 17% when the favourite country chosen is Luxembourg. Moreover, domestic offers that include roaming services have also been launched. For example, in France the postpaid offer Origami Star + Destination Europe (cf Appendix I) grants customer roaming voice services at domestic out of bundle tariff + one hour free to receive calls and direct decrement from domestic bundle. Competitors have also launched innovative offers such as the offer launched end 2010 by Vodafone. Finally and as further described below in our answer to questions 13 and 14, the trend in terms of innovation and decreasing tariffs on roaming markets will be strengthened by the increasing competition on roaming by Wifi, VoIP and soon by LTE. This process of evolution must be duly taken into account by the public Authorities when assessing the current regulation and deciding on any additional one. Consumer protection has been substantially improved In terms of consumer protection, the situation has strongly and positively evolved over the years. After having introduced voluntarily as early as 2008 some protective measures in some of its EU member companies such as Orange France, OFTG has implemented the 2009 regulatory measures, which have further strengthened consumer protection. As noted by BEREC in its reports, European operators have implemented the measures in due time. Consumers are thus better informed of roaming prices and are protected by the anti bill-shock measures. In that context, OFTG has observed a substantial decrease in the number of customer roaming billshocks between 2009 and OFTG has also tried to improve consumers awareness on MB consumption. For example, when describing its Data Daily option on its Internet webpage 7, Orange France also indicates what 10MB means: it allows sending s, or viewing 100 Internet pages, or sending 10 photos or communicating five hours on instant messaging. We will continue to improve customer communications, and provide specific information on the price of roaming. Moreover, in order to better protect occasional roamers from bill-shock, OFTG is activating a default dataroaming package for all new residential customers taking a smartphone. OFTG has also within some of its member companies, such as Orange France or Mobistar, extended the scope of such measures to non EU roaming traffic. It has also granted additional protection or modalities in some countries; for example in France, the fleet managers of 7 nger/etranger_partir.html 7

8 business customers do not only receive SMS but are also called to be alerted when certain thresholds are exceeded. In a recent independent study published by the regulator OFCOM 8, the conclusions on improved awareness are similar overall, there was an improvement in consumer awareness of roaming costs since 2006 : 40% of the respondents indicated that the costs they faced when abroad for roaming calls was broadly what they had expected, 11% considered it was even below. Regarding text, answers are respectively 47% of respondents and 10% and for mobile Internet, answers are respectively 55% and 7%. This confirms that the consumer protection has improved. That study also indicates that 54% of respondents do consider that it has become easier over the last couple of years to find out information about the cost of using mobile phone abroad (while 25% did not have an answer and 15% disagreed) and 46% answered that it is easy to find information on costs of roaming when being abroad (22% did not have an answer and 23% disagreed). Regarding the cut off limit, that study concludes that 56% of respondents were satisfied with it while only 1% answered they were quite dissatisfied and no one claimed to be very dissatisfied. As a conclusion, the current model of regulation has answered the regulatory objectives in terms of consumer protection. OFTG will continue to deliver an ever better protection to its consumers and is introducing additional measures to increase consumers awareness of roaming prices at the time they subscribe to a mobile offer. Regarding roaming prices, the current model has also contributed to price decreases, enhanced by market driven competition, leading to some of the prices being far below the regulated caps, especially when w looking at data wholesale market. 8 International roaming report prepared for OFCOM by Kantar Media, 5 November

9 Question 2: Do you consider that regulatory intervention for roaming services is needed beyond June 2012? Please consider voice, SMS and data roaming services separately. In particular, if you consider that the Roaming Regulation should expire in June 2012, please explain why, and describe how you consider that the market for roaming services will evolve in the absence of regulation. Before deciding to continue with further regulation on roaming markets, and according to the principles of the electronic regulatory framework, it is necessary to show that regulation is required and that wholesale regulation is not sufficient to address the issues at stake; see answer to Q5 below. Any regulatory measures decided must then accord with the objectives and remain proportional. In that sense, while OFTG understands the EC willingness to strengthen the internal market, we also consider that the Digital Agenda for Europe (DAE) target, to reduce the difference between roaming and national charges to close zero by 2015, must be carefully assessed. Such a target, which is not clearly defined (as also highlighted by BEREC), must not set aside the principle behind any regulatory intervention on roaming. That is, to enhance competition to the benefit of end users by preventing excessive prices while taking into account the need for innovation. The need for flexibility is indeed key for operators in order to be able to answer specific consumer expectations; distinctive retail offers are a key factor of competitive mobile markets. The doctrine of proportionality also requires that regulatory intervention should focus on the issue at stake and not impact other competitive markets such as mobile domestic markets. We comment further on this essential issue below. In case the European institutions should decide to continue with regulation on roaming, OFTG requires that any option chosen be fully assessed through a robust impact assessment on which the sector is given a full opportunity to comment. c Such an impact assessment should carefully evaluate all the consequences of the different options on both domestic and roaming markets. Moreover, we have shown above that roaming markets have markedly evolved over the past few years; this has to be taken into account by public Authorities when deciding on such question. This is especially critical for retail roaming data services. As we further illustrate in our answers to following questions, we want to remind the EC that the data market is still a nonmature market, and that, as such, it is undergoing many modifications that the EC should duly take into account when assessing the potential for regulation. We are also confident that both Wifi and new technologies such as LTE will further impact the data market in the coming years. OFTG therefore considers that would the European Institutions decide to extend roaming regulation, they should impose a review by 2014 to take account of ongoing market evolution, especially regarding the data market (see questions 3 and 4). 9

10 II. Options for future regulation Before answering into details to the questions raised in this chapter III, OFTG deems necessary to first highlight some general comments. Domestic benchmark Regarding the options for future regulation, the EC considers it necessary to realise a benchmark of domestic prices in order to verify whether the DAE target is met. To do so, three possibilities are proposed: a benchmark of individual price plan; a national average for domestic prices in an identified relevant market or a European average. Without prejudice to our comments on the DAE target, OFTG considers that obtaining consistent and meaningful results when comparing national mobile domestic prices across the 27 Member States is challenging. While national mobile markets are competitive, domestic prices are and will continue to be different from country to country due to national circumstances and conditions at stake (geography, specific coverage requirements, quality of service, spectrum fees, market demand, prevalence of multi-sim equipment, market segmentation, contract duration etc.). Any domestic price benchmark should thus be analysed and used with the utmost care. To simply average out different countries and/or different user segments risks ignoring national distinctiveness. We therefore consider that such benchmark should be done on individual price plans. EC criteria on effectiveness, efficiency & coherence The EC proposes to assess the future roaming regulatory options in the light of three criteria, effectiveness, efficiency and coherence, which raise the following points. First, OFTG considers it essential that regulation does not lead to costly and/or unfriendly user solutions. While the current roaming system provides quick and easy services to all customers across the EU, we consider that any regulatory obligation that would impose the delivery of a more complex service in terms of billing, numbering or technical implementation for the end users should not be favoured. While the market should remain free to offer such more complex alternatives, actually requiring the players to do so could be highly counterproductive or distorting. We observe that BEREC in its report adopts the same conclusion: whereas roaming prices have been of concern, it has been straightforward for consumers to use the service. Measures which made roaming more complex from a consumer point of viewpoint than the current state of affairs therefore seem unlikely to be favoured 9. OFTG then welcomes the fact that the EC deems it necessary to take into account ease of use, administrative and compliance costs for the operators, expected loss in profits and technical complexity. This should clearly be part of the impact assessment the EC will have to deliver before proposing any further roaming regulation. OFTG is at the EC disposal to contribute to such analysis. For example, technical feasibility, and corresponding implementation costs, are key when assessing an option; technical development imposed by regulation can deter innovation and mobile operators do not have unlimited resources dedicated to roaming, let alone the implementation of new roaming regulation. As an example, while we support the anti bill-shock measures imposed by the 2009 regulation as a matter of principle, their concrete implementation in a very short timeline has been quite difficult and burdensome in particular for smaller operators. Such reality must be part of the discussion, as also indicated by BEREC in its aforementioned report, BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, 56 10

11 We observe that under the effectiveness criteria, regulatory options should be assessed in light of achieving the domestic prices target of the DAE; while under the coherence criteria, any solution should avoid market fragmentation and distortions of competition. We wonder whether both criteria are achievable at the same time. As indicated by BEREC in its report, the DAE target could create concerns 10. Further clarification from the EC on how to conciliate both objectives appears therefore necessary. Proportionality: key principle of the regulatory framework Finally, we consider that the EC should also remember that any regulatory intervention on a market has to remain proportionate: the means to achieve a legitimate aim must be both necessary and the least burdensome; i.e. it must be the minimum necessary to achieve that specific aim. This principle of proportionality is a pillar of the electronic communication framework and should be further taken into account by the EC in its criteria when assessing the different options quoted in the consultation. We therefore request the EC to add in its list a criteria based on proportionality. Proportionality also means that regulatory measures should be sufficiently balanced to achieve diverse objectives of regulation such as competition, innovation, incentive to invest, etc. As part of the proportionality assessment, the potential impact of roaming regulation on national markets must be taken into account. A roaming regulation cannot be the way to regulate mobile domestic markets, such as mobile domestic access, that are competitive and not concerned by potential roaming competitive issues. This would distort competition and be in breach of both the principle of proportionality and market analysis process defined by the electronic communication framework. In that sense, we observe that BEREC, in its list of criteria, does include such potential impacts on national markets. In addition, we also consider that scenarios that would impose both a structural option and a price-cap would be highly questionable in terms of proportionality. 10 BEREC aforementioned report, 101 : «However, if EDA target is expressed in terms of domestic prices paid by individual customers, it is not an ideal long term solution as it tends to lead to near uniform prices. See also 119 according to which a strict interpretation of the EDA target (i.e. roaming prices extremely close to domestic prices) could create concerns. There is a real danger that some providers would be required to sell roaming services at level that do not allow them to earn an adequate return on their activities. 11

12 a) Maintain current approach Question 3: Do you consider that the current model of regulation would be effective in the future in light of the desired objectives? Will this approach ensure adequate consumer protection and help stimulate competition? Is it efficient and coherent with EU policies? Question 4: If this model is suitable in principle, what modifications may be required in order to achieve a well functioning single market for roaming services? Should this approach be combined with other options? As indicated in answer to questions 1 and 2, the EU roaming market has strongly evolved since the first regulation entered into force both in terms of price levels and consumer protection. There is no reason to believe that such evolution would not continue to develop positively in the event the EC wants to continue with the current type of regulation. Current approach & consumer protection In terms of consumer protection, we have indicated above how the current model has showed positive outcome. However, OFTG is committed to deliver an ever better protection to its consumers and has already gone further than regulated solution. OFTG is currently thinking about introducing additional measures to increase consumers awareness of roaming prices at the time they subscribe to a mobile offer. OFTG is activating by default a data roaming bundle for all new residential customers taking a smartphone. We therefore consider that the current model of regulation has answered the regulatory objectives in terms of consumer protection. We do not consider it necessary or justified to impose further transparency measures; transparency requirements are now well established within the sector and a are even further improved by some operators like OFTG. However, implementation of the anti bill-shock measures to certain type of traffic has raised some questions and difficulties that could be clarified in a new version of the regulation. The 2009 regulation has indeed not taken into account some specific offers that match with difficulty with the anti bill-shock measures. As also underlined by BEREC 11, this is the case for prepaid offers as well as MMS. Neither were thought of in the 2009 regulation and thus they were a source of confusion, to the point that ERG was required to issue specific guidelines for MMS 12. This still leaves mobile operators in a legal void, as these guidelines do not supersede the roaming regulation. Similarly, we wish to point out that M2M traffic ought to be removed from the scope of those transparency measures, given the specificities 13 of such traffic. Current approach and price regulation In terms of roaming price regulation, should the European institutions decide that additional regulation is required after 2012, OFTG considers that the current approach - based on a glide path,, imposing reasonable and proportionate decreases and acknowledging data market distinctiveness - should be favoured over other alternatives for r several reasons. First, as elaborated below, after internal analysis of the diverse alternative solutions proposed by the EC in its consultation, none of them lead to a satisfactory situation, notably especially in terms of efficiency, costs or ease of use for the end users. Second, any further roaming 11 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, p164 to International Roaming Regulation ERG Guidelines, ERG (09) 24 final, See page 17 12

13 regulation should be compliant with the regulatory framework, according to which remedies have to be proportionate, transparent, focused on the issues at stake (i.e. roaming), avoid any damaging side effects on other markets not subject to regulation. Thirdly, a glide path with reasonable decreases would ensure a better predictability and would be easier to implement. Such an approach would also allow some space for innovation and retail segmentation. A regulation based on a glide path therefore appears to be more in line with the objectives of the EU electronic communications framework, as long as specificities of data market are taken into account. Our internal assessment leads to a conclusion similar to the one adopted by BEREC in its report according to which: no option is definitively superior to the approach which has been followed so far and the alternatives all have downside risks 14. We therefore consider that should additional regulation be decided after 2012, it should have the form of a glide path with reasonable and proportionate decreases. In any event, any new regulation should strictly avoid regulation of domestic mobile access, as it would be highly disproportionate and contrary to the electronic communications framework. We refer for more details on this essential topic to our answer to Q12. An in-depth impact assessment of the burden and feasibility of any regulatory measures proposed by the EC, including impact on domestic markets, should be delivered sufficiently in advance in order for the sector to have the possibility to assess and discuss it. Preliminary remarks on any new future glide path In case the EC would propose new price caps for future regulation, we consider that it should distinguish between voice, SMS and data. Regarding voice, SMS and wholesale data, any new price caps should be fixed to reflect current market and regulatory conditions in the mobile sector while not impacting domestic mobile prices. It should take into account mobile operators costs to provide roaming services across Europe while giving sufficient space for innovation and retail offers segmentation. As far as data retail is concerned, we would favour a cautious approach from the EC in order not to distort the ongoing competition market led and innovation. As further explained under questions 6 & 7, the data market is still not stabilised or mature in terms of business model and offers at domestic levels, meaning it is even less the case at roaming level. However, prices have been decreasing and competition is playing and will be strengthened in particular by domestic dongles or by other technologies such as Wifi. Finally, LTE might also deliver new roaming solutions. Any new regulation on such evolving and dynamic market could have the counterproductive effects to distort the current trend of market led competition (see also Q6&7). Regarding BEREC s estimation made for 2012 wholesale and retail costs to provide roaming voice, text and data, OFTG is at BEREC disposal to further discuss that issue but already wishes to stress that: The data sample is at this stage quite limited; cost assessment has been made based on the data from 9 NRAs, which could bias the results of the study since costs can differ importantly from one country to another. OFTG urges for the utmost caution if this study is to be used as a start point to work on price levels. BEREC s report doesn t mention to what extent coverage costs were included; 14 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, 55 13

14 The model doesn t take into account a country per country segmentation, and thus leads to a one size fits all pricing across Europe, detrimental both to the industry but also to customers The retail model is based on the wholesale cost and not on the wholesale price, which is economically questionable as it assumes that the wholesale operator doesn t make any profits in selling wholesale traffic. Finally, regarding its validity period, we consider that a review of any new regulation should be done by 2014, as proposed by BEREC. By that date, the roaming markets will have further evolved due to data roaming growth driving new uses and services in parallel to the further development of data roaming substitutes such as Wifi. By 2014, all stakeholders will thus be in a position to better assess the consequences of the data roaming growth on roaming markets and competition. In conclusion, on questions 3 & 4, 4 OFTG considers that should the European institutions decide to continue with roaming regulation: Additional regulatory transparency measures es are not required; o Transparency measures are now well established within the sector and benefit all end users. Strengthening or improving such measures is now part of the operators differentiation strategy. Additional compulsory measures on transparency would therefore not be justified; o Some clarification could however be introduced regarding M2M, prepaid or MMS services. The form of a price cap should be preferred to any other alternative solutions,, which all have substantial downside risks or do not deliver the same customer experience.. Such price caps should meet the following requirements: o Distinct treatment for voice, SMS and data; to take into account the non mature aspect of the data market; any data retail regulation to be carefully assessed; o Should a price decrease be chosen, the glide g path should be proportionate and reasonable; while taking into account all impacts of such regulation on roaming and domestic markets on a consumer and operators point of view; a new tariff decrease must avoid any unintended consequences on innovation and on current domestic retail markets that are competitive and characterised by offers segmentation; BEREC methodology and costs c estimation is to be further discussed with stakeholders o Impact assessment to be carefully done on roaming and domestic markets, and discuss with stakeholders. A review r should be planned for 2014,, to take into account data market evolutions and ongoing market led competition. 14

15 b) Wholesale and transparency measures only Question 5: Would regulation of wholesale prices charged to MNOs, combined with transparency measures, be effective, efficient and coherent in light of the single market objective? Would the benefits of regulated wholesale rates be passed through to consumers? OFTG considers that as a matter of principle, and in line with the framework, should there be a market failure, then regulators should first address the wholesale market and if it remains insufficient, then the need for retail market regulation can be assessed. Roaming markets have shown strong dynamism particularly when looking at wholesale data roaming. European institutions should ensure, when opting for regulation, that the type of measures they adopt will not stop or delay market led dynamism. If European institutions consider it necessary to intervene on roaming markets, wholesale regulation in the form of a glide path together with transparency measures should deliver positive outcomes. These measures would be efficient in a context of important smartphone investment from operators, increasing customer awareness of cost and alternative solution over Wifi. In that sense, we do not agree with the EC concern expressed in this section according to which there would be a risk of retail price increase if roaming regulation were only to focus on the wholesale market. Considering the market forces at stake on roaming markets together with the new roaming offers being launched, and mobile communication prices being in a downwards spiral as a result of fierce domestic competition, OFTG strongly believes that it would simply be impossible to increase roaming charges, even if the current retail price controls were lifted. It has also to be noted that any retail price increase provides a right for termination of the contract by the end user; mobile operators are not going to take the risk of loosing their clients in such a competitive market by increasing retail roaming prices. The option to raise prices if the roaming regulation at retail level is not extended is thus purely theoretical. As a consequence, we consider that the risk of increased retail prices, if roaming retail regulation were lifted, is non-existent in the mobile context. As far as the pass-through to consumers is concerned, the retail roaming market is evolving with the launch of new attractive offers; see Appendix 1. The explosion of data usage will also substantially affect the functioning of roaming markets: this trend was initially a domestic one, but our customers expectations are now more and more to benefit from similar services when roaming, thus increasing the overall competitiveness of the retail data market. We therefore consider that roaming conditions will continue to evolve towards greater competition. Regarding transparency to end users; current measures have been shown to be an effective tool to improve consumers awareness on tariffs and increase trust and confidence in roaming services; see also comments made above. OFTG anticipated, in some of its EU footprint countries, the transparency and anti bill-shock measures imposed by the roaming II regulation. Since it has entered into force, OFTG has been compliant with such measures that, even if quite complex to implement for smaller players, have proven to be useful in terms of consumers protection. OFTG has also extended the scope of such measures to non EU roaming traffic in some of its member companies. As a conclusion, proportionate and reasonable wholesale prices regulation, combined with transparency measures, should be the first answer r to any competitive issue on roaming markets. 15

16 c) Regulation of retail data roaming charges Question 6: Do you consider that retail regulation of data roaming prices is necessary? If not, what are the likely market developments post-june 2012? Question 7: If retail regulation of data roaming prices was necessary, what would be an appropriate model for such regulation? Data retail market characteristics: growth, innovation, price decreases, segmentation and lack of maturity Mobile broadband is characterised by a tremendous growth: in a recent study from AT Kearney 15, it has been estimated that mobile broadband will increase by up to 100% per year till The launch of innovative smartphones is clearly a factor enhancing such trend; we observe that over 50% of all new mobile phones acquired in Western Europe in 2010 were smartphones, while there were only 5% of them two years ago: a tenfold increase of smartphone penetration. Innovation and quick market evolution characterise this market, notably with smartphone producers continuing to regularly improve capacity of their products. The recent launch of tablets is another illustration. It has been a success from the start, modifying again consumers expectation and consumption. For the time being, in the face of such traffic increases, business models are evolving. Mobile operators are still regularly modifying their retail data pricing models in the search for one that corresponds to markets needs while remaining sustainable in the long run. Offers based on time use, downloaded volumes, numbers of connections, as well as unlimited data offers, have all been tried, without any of them fully meeting the market s needs. This lack of maturity in terms of business model is present worldwide. For example in Europe, Orange France is reviewing its initial strategy to launch unlimited data offers. T Mobile in the UK has also decided recently to review its strategy to limit the use of data. Similarly, in the USA, Verizon has decided to adapt its initial strategy and no longer grants unlimited data use. Domestic data markets are at the same time characterised by aggressive retail offers within bundles to attract customers and develop usage of such new products. Such a competitive trend is now passing on roaming markets. For example since February 2010, OFTG has developed monthly bundles for different customers segments as well as attractive opt-in daily offers Data Daily offer as further described in Appendix I (2 /2MB on mobile, 5 /10MB on smartphone or 15 /50MB on tablets & notebooks) that provide retail tariffs below the average of /MB 16 as identified by BEREC in Q This offer has already convinced about 1 Million customers so far. OFTG has also started to offer per default to all its new residential customers buying a smartphone a specific roaming bundle similar to the Data Daily offer. The business segment is also very dynamic; for instance a Mobistar retail data offer, "Roaming Internet proposes since 2008 retail data prices especially low: 500 MB for 95 meaning roamers pay 0.19 /MB while the wholesale regulated cap is of 0.80 /MB since 1July While this offer only generates 25% of the total data roaming revenue, it represents 40% of total data roaming volume. Similarly in France, Orange has launched in February 2011 a new offer for 15 AT Kearney study on the Viable Future Model for the Internet ; 16 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, Table 1, p6, Average retail price for data, including prepaid and postpaid, on-net and off-net 16

17 business (including SME): Europe Ajustable offers, for a 10 monthly fee, from 5 MB to 3 GB data roaming with decreasing prices depending on the volume, automatically adjusting itself to the cheapest data roaming plan. With this offer, any business user can benefit from a retail price as low as 0.45 /MB, whenever he uses at least 100MB of data roaming in the month. Competitors like Vodafone have also launched aggressive retail offers late in These retail offers aim to answer diverse and specific consumers needs. The data market covers different types of traffic and products. Video traffic for example differs from , that in turn differs from M2M traffic. Some services can be more quality of service dependent than others, which could require different offers. Surfing via dongles or mobile phone or tablets requires different characteristics that can justify different type of offer. M2M solutions, integrated for instance in cars and electric boards, are very specific as they generate only a very small amount of traffic, as little as 1MB per month, while in the same time remaining continuously on and thus generating significant (and free) signalling traffic. Data retail offers segmentation is therefore essential to match the diversity of data consumption. In its report, BEREC acknowledged that it is, aware of several recent commercial price initiatives on data roaming which offer consumers in some countries very substantial price reductions. These may be followed by competitive responses over the coming months 17. In its regular report on roaming, BEREC also observed a clear trend of data retail prices decrease: The picture is different for data roaming, in that the average wholesale and retail charges have fallen significantly between the second quarter of 2007 and the second quarter of As we pointed out in question 1, this statement is backed by BEREC benchmark figures, which showed an average retail data price in Europe of 5.81 /MB in Q dropping down to /MB in Q , representing nearly a 78% drop in the retail price of roaming data in three years. Here again, OFTG underlines that this substantial drop was done without any price regulation of retail data roaming. Price decreases and transparency measures that give consumers a better awareness in terms of prices favour the popularisation of data usage. Within some of OFTG member companies, customer data roaming traffic rose by about 50% in 2010 compared to At the same time, OFTG indeed observes a clear decrease of data bill-shock. For instance, within Mobistar, complaints relating to roaming bill-shock decreased very significantly between 2009 and 2010: level of 2010 being around 1/3 rd of the 2009 level 20. Such a trend is strengthened by alternative competitive solutions such as national dongles, Wifi or new technologies such as LTE, that will further impact the data market in the coming years (cf Q13 & Q14). In that sense, it is worth highlighting that in 2009 the European institutions decided not to regulate retail data roaming market because competition constraints were at play: unlike voice and SMS roaming services, competitive constraints exist at retail level, as roaming customers have alternative means of accessing data services when abroad, such as public wireless access to the Internet 21. Such features and conclusions remain valid at this day given the fast evolution of data usage. It is also worth reminding that BEREC in its report indicates that in the longer term, retail data roaming could become a reasonably competitive market, given that there are partial substitutes (e.g. via Wifi access), which provide competitive pressure Aforementioned BEREC report 108 p International Roaming BEREC Benchmark Data Report for January 2010 June 2010, p4. 19 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, Table 1, p6, Average retail price for data, including prepaid and postpaid, on-net and off-net 20 [CONFIDENTIAL] Roaming Regulation, recital BEREC report on Roaming, 41 p 11 17

18 Careful assessment required before any data retail regulation Taking into account the above, we consider that the retail data market is still far from being mature while facing increasing competition pressure and better consumer awareness and protection. All those specificities attached to the data retail, domestic & roaming, markets should lead the public authorities to adopt a particularly pragmatic and prudent approach when assessing the need of data retail regulation. When deciding on any potential data retail regulation, the institutions should ensure that such regulation would: Not distort market led competition: competition is developing and increasing within data retail market. This trend will further increase with competition by domestic dongles, Wifi or in the future LTE. Any regulation should thus be particularly cautious on the disruptive impact it could have on this positive market led trend. Not distort domestic markets: o As indicated above, the domestic markets themselves are still subject to constant evolution and innovation in the search for sustainable data offers in line with customer expectations. Basing any data roaming retail regulation on domestic data markets would therefore be highly risky since current evolving domestic bundles can lead to an incorrect perception in terms of costs and business models. o Domestic markets are competitive while varying strongly among Member States; any retail data roaming regulation should ensure that none of the country would be squeezed by this regulation. Avoid distortion ion of innovation and domestic offer o segmentation: domestic markets are competitive and retail offers answer different customer needs, from low consumers to business consumers via M2M users etc. Flexibility to adapt retail offers to specific needs is key. Any regulation that would affect such flexibility and that would not allow adaptation to low or high volume session would be highly questionable. Carry out a robust impact i assessment: the EC is considering different options for retail data charges: a Euro data tariff or an average charge per MB. In all cases, any data retail charge regulation would require an in-depth discussion between stakeholders, BEREC and the European Commission in order to ensure a common understanding of costs, market dynamism and impact on national and roaming markets. Review by 2014 to practically assess the impact of the new regulation as well as new market developments. As indicated above, that market is particularly dynamic: its competitive evolution should not be hampered by out-of-date regulation. 18

19 d) Approaches based on prices and conditions similar to those prevailing in the domestic markets Question 8: Please indicate the advantages and disadvantages of these approaches, relative to each other and to the current model of price capping, considering also competition aspects such as the possibility of margin squeeze? This is the scenario based on the DAE objective defined by the EC. In its consultation, the EC indicates that it could mean to set a roaming price close to domestic home prices (with a reasonable mark up) or to domestic visited network by While understanding the EC wish to enhance the Internal Market, we consider that this target carries a high risk of distorting the mobile domestic markets and thus raises a question of proportionality. First of all, the concept of domestic prices is difficult to define. There is no unique price and having to define one to fix roaming prices raises serious concerns. As also acknowledged by BEREC in its report, domestic markets naturally present distinct features (geography, cost of spectrum, coverage or quality of services requirements, market preferences) which directly impact the level of prices paid by consumers. The nationally defined framework for the mobile industry explains in great part price differences amongst Member States. In 2011 and 2012, we can expect these national differences to be exacerbated with the spectrum auctions announced in a large number of member States. Competition in those national markets will not and cannot itself deliver, in the short run, equal prices across the European Union. We therefore do not see clearly how the European institutions could impose a unique reference price for roaming services, or why it would try to do so. Moreover, the domestic mobile markets have been and are competitive (innovation, penetration, coverage, unit prices. ). One reason for this success is the segmentation of the offers which meets customer needs but inherently introduces a broad range of price tags for the different components of the offers. Any roaming regulation that would put at risk such segmentation by imposing one type of offer or level would be unacceptable. In the advent of roaming prices and conditions forcefully made similar to those prevailing in the domestic markets, OFTG is particularly concerned about the spread of the spill-over of Roaming Regulation to a much more significant part of all domestic markets. In that sense, we fully agree with the criticisms made by BEREC when it indicates that definition of such price would be highly difficult 23 and that, fairly wide variations are normal, even in a competitive market with no great variations in the costs of provision. Where efficiently incurred costs are different, uniform pricing leads to distortion or discrimination between different players. For example, if implemented in conjunction with a requirement that the roaming price equals the European average domestic price, roughly half of European consumers would pay less for roaming services than for domestic services. This seems a rather bizarre outcome 24. Roaming regulation shall not impact domestic competitive markets; this would go far beyond the alleged roaming competitive issues at stake and would thus be in breach of the European framework, notably the principle of proportionality. The financial impact for the mobile sector to achieve such target by 2015 would moreover be tremendous, raising again questions in terms of proportionality. Such impact could also render more difficult other objectives defined within aforementioned BEREC report on Roaming, p aforementioned BEREC report on Roaming, p 13 19

20 the DAE and notably those on fast and very fast broadband roll-out. Finally, achieving such target at retail level could lead to a non-sustainable situation because of squeeze effect. Moreover, when looking at customer transparency, having roaming prices based on visited network prices would raise numerous difficulties: what would be the price level defined as the domestic level? How will the customer be informed? By who? Furthermore, it is unclear how such a solution could be practical when the customer s home country doesn t have the same currency as the visited country. How could a customer be made aware of the price he or she would pay with ever-changing exchange rates? Regardless of the actual definition of the domestic price reference, asserting that roaming should be priced as a domestic communication fails to account for the many specificities of roaming traffic. Domestic users will tend to use their phones the same way regardless of where they are and what time in the year it is. A domestic user will also tend to use mainly on-net communications. On the other hand, a roaming customer will generate nearly exclusively international communications, at very specific times in the year and in particular places. These peaks of consumption are such that they affect the design of mobile networks in some popular tourist areas 25. Furthermore, different specific cost elements 26 would need to be accounted for in a mobile user s domestic pricing, as they are specific to roaming. We also wish to underline one of BEREC s points from its report 27 : There is a risk of arbitrage between offers from different countries, which would make it difficult to sustain the business model [of handset subsidies]. OFTG feels very strongly against such pressure against a domestic business model handset subsidies in this case - that is widely used within the industry, especially as it allows us to facilitate the adoption of smartphones and similar datacompliant handsets. As a conclusion, we consider that current roaming markets price, in particular data roaming, are showing a clear trend of price decreases. Imposing now that by 2015 roaming prices have to be close to domestic prices, would appear disproportionate and would highly risk having counterproductive effects at national level, where markets are competitive. 25 This was duly noted in recital 17 of the Roaming Regulation (EC) N 544/2009: Some operators face higher wholesale costs than others due to geographical or other circumstances, such as difficult topography, regions with low population density and large influxes of tourists within short time periods. 26 Such as the termination cost, commercial costs for all the roaming advertisement in areas receiving tourists, testing and billing costs, 27 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, p112 20

21 e) Separate sale of roaming services - decoupling of roaming from mobile services s bundles Question 9: In general, would these decoupling approaches be effective in terms of stimulating greater competition for roaming services? Would all customer segments be able to benefit? Would such increased competition be sufficient to give consumers an effective choice of roaming services at (near) domestic prices? Question 10: Would such 'structural' approaches be efficient? What are the technical implementation issues associated with these approaches? Carrier Pre-Select in the visited network and Choose operator at the border Both the CPS in the visited network and the Choose operator at the border scenarios would be a source of many problems, impacting customer friendliness and transparency, while imposing a significant burden on operators without any obvious impact on competition. Based on our knowledge of the mobile market, we feel that requiring that users choose between unknown providers in the visited country would be stressful for many of the occasional roamers. In addition, the roamer would experience the same network coverage as would a domestic user of that visited network. This would be detrimental to the user s ability to use his mobile phone: If switching to another network were to be only done by the user, these coverage problems would occur in areas in which, by definition, there would be no available information (regarding tariff, how to access the terminal s menu to select provider ) helping the user to select another provider. Many users are not familiar with the process of selecting a network. In fact, on some handsets, network selection menu can be very difficult to find. If automatic switching were to be authorised, end-users could end up on another network than expected, and pay another rate, in all cases causing unexpected bills and potentially even leading to bill-shock. The loss of control in the steering of traffic could be detrimental to user s perceived quality: mobile users aren t capable of measuring a visited network s quality (audio quality, bad coverage, interrupted calls, technical compatibility of services) as home providers do today. OFTG also wants to point out that such a solution would also be detrimental to smaller providers in their ability to compete on roaming: whereas, today, preferring a smaller network as VPLMN is possible as this selection could still fallback on a wider network if only this wider network is available. In this scenario roamers wouldn t be enticed to choose these smaller players if there wouldn t be any automatic switch to an alternative available network with a wider coverage. The steering ban for Europeans in Europe that derives from this proposal would need to be taken into account. It could affect the roaming usage of Europeans roaming in the RoW because the habit of having to select VPLMN wouldn t apply outside Europe. This scenario also introduces numerous sources of complexity on the mobile operator s side: As prepaid subscribers are billed based on Home SCP, a complex tariff plan synchronization mechanism would need to be put in place. The efforts to achieve such a result appear to be significant and would ultimately cause errors and additional cost (as also highlighted by BEREC p BEREC on p102 of its Roaming Report). The liabilities of both providers also unclear: for instance, if the visited network were to incorrectly price a tariff, who would be liable? If the customer complains about a tariff, which operator has to deal with such issue? More generally, we are unclear of what would be the legal binding between the Visited Network and the mobile user. 21

22 Moreover, such a solution would also require each operator in the EU to interconnect his billing information system with those of all operators in the 26 other countries, and ensure a full compatibility. The previous regulatory requirements regarding anti bill-shock cut-off would be even more complex to attain, as well as those regarding communication on the prices due to the enormous combination of possible billing plans, over all countries and all providers. The VPMN would have to present his retail price in local currency with local VAT. The end user would then be facing issues regarding exchange rate fluctuations and VAT (for business customers). Operators would have difficulty informing consumers of the available offers: SMS don t contain enough characters, and free phone numbers or website would need to be translated in at least all the 23 official EU languages. Keeping up to date the pricing information in all 23 languages would prove to be a nightmare for operators. We share BEREC s view 28 that this would also require additional measures to regulate any information services provided by the home operator to avoid any commercial bias towards its own products. Finally, we also share the EC s concern regarding the level of competition that would derive from either of these two scenarios. Both assume that there is in each country a market for roaming products alone. This would need to be demonstrated, and it seems at the very least a hasty assumption. Such competition would be solely driven by tourism, and Eurostat s latest tourism statistics 29 illustrates a significant heterogeneity between tourist flows, to the point that mobile operators in less visited countries would have little if any incentives to promote competition, or even create such offers. Carrier Pre-Select in the domestic market This scenario is slightly different from the two previous ones. Though it solves some of the issues raised above, other problems are created. As it is pointed out in the public consultation, the flows of calls themselves, of billing and of signalling would need to be significantly modified. Currently, the HLR is the only equipment that stores the customers details. Having to duplicate it to another provider or having some of its functions done by any other equipment, especially one that belongs to a competing company is rather questionable. None of the modifications are accounted for in the roaming standards, which will lead to different implementations. At best, these implementations will be consistent on a national level, but it is quite plausible that incompatibilities remain even between national operators that are supposed to interconnect following this new requirement. Furthermore, as BEREC stresses that this is not a standard procedure, significant implementation will be required, backed by standardisation activities 30. OFTG agrees with that statement, underlining the key issues regarding security and privacy if single IMSI SIM cards were to be used. Nothing in the current standards is planned to share authentication and registration information as is suggested in this approach. In all cases, OFTG wants to make clear that customers security and privacy must not be compromised in any way by the roaming regulation, and that it would take time to find manageable solutions to such issues and that no implementation could be expected in a short time. As BEREC points out, the use of multiple IMSI SIM cards would reduce the extent of the problems previously listed, but not remove them completely. Due to compatibility issues, 28 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, p BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, p110 22

23 requiring multiple IMSI SIM cards would require most users to change both handsets and SIM card. Such a process would be extremely costly for operators as well as for users. We are also concerned about the liabilities within this CPS approach: customers wishing to complain about their roaming bills would surely contact their provider to which they subscribe for their main contract. This system might render such process more complex for the end users. The same would apply to any customer that is roaming and who would want to contact his customer support: depending of his problem, the customer would have to call either his domestic provider or his roaming provider. As a conclusion, the option CPS in domestic market would be complex for end users and operators, and would require a reasonable period of time before any implementation. The situation is even more complex for the two other options (CPS visited network & choose operator at the border). We also consider that those options could be useful only if roamers find it attractive and easy to handle, which as shown above is doubtful especially for the two first options analysed. Finally, they would also depend on CPS roaming consumers willingness or interest to launch campaigns to be known as roaming providers and attractive roaming offers Ther here are currently many more questions than there are ready answers regarding CPS solutions. We consider that such options, especially the first two mentioned above, would be significantly less user-friendly friendly,, very complex also for the operators to implement and with uncertain outcomes at retail level by CPS roaming providers. In any cases, were the EC interested in supporting one of these CPS options, a serious study on its impact and on its implementation would be required, especially if it were to be combined with options f) or g), which raise other concerns (see following answers). 23

24 f) Spot-trading trading of wholesale roaming Question 11: How feasible/efficient is the establishment of a spot trading market for wholesale roaming? Would this approach lead to competitive wholesale rates? How effective would this approach be in terms of achieving competitive retail rates? OFTG is strongly opposed to the idea of establishing a spot trading market for wholesale roaming. In the light of the considerable changes involved by this solution and its impact on all parties, OFTG is very concerned that such a proposal could be suggested without any prior demonstration of its expected positive outcomes, let alone the lack of any in-depth assessment of its impacts. In our view, this scenario represents a significant risk: nobody has a clear understanding of how this solution would be implemented in the mobile environment. At the very least, spot-trading would be challenging from a technical perspective and from a billing/fraud perspective. On the technical side, as no standard has been made regarding such a solution, there is a significant risk that different players of the European mobile market develop and deploy different noncompatible solutions to this new requirement, leading ultimately to even greater costs and potentially some bilateral routes within Europe having to close. Moreover, as also indicated by BEREC in its report, roaming cannot be regarded as a standardised commodity. Coverage and quality vary from buyer to buyer. On a spot trading market requiring blind trade, roaming services would need to be presented in a standardized way, which would impede tailor-made agreements. On the other hand, the anonymity of buyers and sellers might soon or later be quickly discovered. A pure spot trading market might lead to some damaging effects where operators based in receiving countries might take benefit from the higher demand for roaming. These operators could decide, instead of raising network capacities in time of high demand, to limit or keep saturated network capacities and raise prices. On the billing/fraud side, as mobile operators participating to the spot-trading would be bidding anonymously, we are very concerned about not being able to check the credentials and the credit rating of the other bidder, prior to closing a deal. This will represent an additional risk, which will need to be accounted for financially. More generally, a new kind of regulation would occupy the mobile industry s focus, instead of working on innovative new products. In its consultation, the EC is rather vague about the actual details of such a solution. If it were to follow the route discussed by some 31, it would raise serious concerns as explained by BEREC in its report 32. Such approach would indeed not take into account structural asymmetries between countries exchanging traffic, and the fact that competition is strong particularly on unbalanced traffic. As is further explained in questions 21-22, what is important in wholesale roaming negotiations is more the steering capabilities, coverage and quality rather than just the operator size. The analysis also leaves aside all considerations regarding network capacities and the existence of an opportunity cost in the usage of mobile networks. Either the volumes proposed on the market will be very limited; either MNO will be taking risks regarding the networks capacities to deal with all the traffic. Therefore, the way for this analysis to apply would be by dumping a network s quality and its future. All of this also raises questions regarding the proportionality of the measure. Finally, we have concerns that this solution could also introduce speculation issues. 31 Tony Shortall, A structural Solution to Roaming in Europe, EUI Working Paper RSCAS 2010/62 ( 32 BEREC report on International Mobile Roaming Regulation, December 2010, BoR (10) 58, p

25 g) Access- based approaches Wholesale retail for MVNO (roaming and all mobile services) Question 12: For each of options (i) to (iii) above please indicate whether such approaches can stimulate additional competition for roaming services. In order to achieve significant reductions in roaming prices do you consider that these 'access-based' approaches may need to be combined with other forms of wholesale price regulation (i.e. between MNOs) and/or retail price regulation? Please explain. The 3 options 33 proposed by the EC in this question aim to impose to MNOs a regulated wholesale access at the benefit of MVNOs regarding roaming wholesale prices only or even for all mobile services generally. Such proposals raise serious concerns in particular regarding (iii) as further explained below. MVNO agreements are freely negotiated on domestic competitive market Those options seem to be based on a perception according to which an MVNO would not be granted relevant or satisfactory conditions at wholesale level. This is not correct. MVNO agreements are negotiated freely between MNOs and MVNOs. Negotiation on MVNO agreements lead to a fierce competition between MNOs at domestic level, giving MVNOs the possibilities to negotiate interesting conditions, including for roaming if they wish so. Interesting to note in that sense that in one OFTG country footprint one MVNO is hosted by different operators according to the services purchased: one MNO is providing the domestic services and another is providing the roaming services. Similarly at European level, companies like Transatel do negotiate wholesale roaming deals with different domestic MNOs. It can also be observed that the current competitive situation at the wholesale level does not necessarily result from the MVNOs lack of access to the wholesale roaming market, but from the MVNOs policy. MVNOs very often prefer to avoid entering into time consuming roaming negotiations which imply labour and management costs while the expected revenues might be low. We therefore question the necessity or proportionality of any obligation to deliver wholesale regulated access to MVNOs for roaming services (options i & ii). The situation is even more questionable when looking at option (iii) according to which such wholesale access could cover not only roaming but also domestic services. Regulation of MVNO agreements: no justification On a legal point of view, the three proposals made by the EC raise several questions regarding their compliancy with the regulatory framework, especially when looking at option (iii). Imposing a regulated wholesale access for MVNO would lead to the regulation of ex market 15 of the 1 st EC recommendation on the relevant markets adopted in At that time, the EC deemed necessary to request NRA to assess the competition conditions of this market. Four years later, the EC decided to review this recommendation and to withdraw several markets among which ex market 15. As from 2007, NRAs therefore do not have any obligation to launch market analysis on such market. Such decision has been taken by the EC in the light of market 33 (i) regulated roaming wholesale access, (ii) wholesale cap for MVNO, (iii) regulated wholesale access for all services 25

26 analysis done by the NRA concluding in a vast majority of States that ex market 15 was fully competitive; see Appendix 2 showing the EC table on market analysis for Cocom meetings. In its explanatory note to the EC recommendation dated 2007, the EC clearly states that (i) mobile retail markets are competitive across the EU and (ii) wholesale access and call origination mobile markets are also competitive at EU level: the degree of competition generally observed in this market [mobile market] at the retail level indicates that ex ante regulatory intervention at a wholesale level may not be warranted. In addition, in most Member States the wholesale mobile access and call origination market is effectively competitive as mobile network operators conclude access agreements on commercial terms. [ ] there have been cases where there have been findings of single or joint dominance. However, there is no evidence to suggest that action under competition law could not address such problems. In fact it may be that competition law instruments are particularly well suited to addressing the issues that may arise. The market is therefore removed from the recommended list 34. This scenario proposed by the EC, and especially option (iii), appear thus seriously questionable, as it would contravene market analysis process (requiring market definition, SMP designation and then remedy), NRA competencies (sole competent to intervene at domestic level) and contradict NRA as well as EC conclusions on ex market 15. Any roaming regulation imposing such obligation, and more obviously option (iii), would thus breach the electronic communication framework and must therefore be avoided. BEREC does not say otherwise when indicating that, the measure amounts, more or less, to cost oriented regulation of the old market 15 without finding of Significant Market Power. Some believe this could raise significant legal issues 35. Moreover, this scenario g) raises also questions in terms of proportionality. As reminded above on p 11, one of the pillars of the regulatory framework is the principle of proportionality according to which regulatory measures must not go beyond what is necessary to achieve legitimate goals. The objective of the regulation on roaming is to tackle alleged excessive roaming prices; any regulatory measures should thus focus on that objective and avoid impacting other markets like mobile domestic markets. In that sense, regarding options (i) & (ii), the EC should further describe how such access could be practically limited to roaming services without impacting domestic competition and terms of MVNO contracts. There is a serious risk, and more especially option (iii), to directly impact wholesale domestic negotiations between MNOs and MVNOs, including prices. Option (iii), options (i) & (ii) in a lesser extent, would indeed grant a universal right to become MVNO on regulated terms, whose prices may impact domestic prices. There is thus a serious risk that such regulation would impact beyond roaming services and put at risk competitive and freely negotiated MVNO agreements signed in the Member States as well as competition conditions at stake at national level. Such regulated wholesale access for roaming and possibly domestic markets would moreover disregard MNO investments and specific constraints imposed to them. Imposing an unjustified regulated access for MVNOs via roaming regulation would render negotiation unbalanced between MVNOs and MNOs, which are submitted, among other things, to license obligations, network related costs and spectrum fees. This could also question the interest of MNO to further invest in spectrum, meaning that such regulation could also negatively affect investments and technology developments. 34 Explanatory note to the EC recommendation on relevant product and service markets within the electronic communications sector susceptible to ex ante regulation in accordance with directive 2002/21/EC of the European Parliament and the Council on a common regulatory framework for electronic communication networks and services, p BEREC aforementioned report on roaming, 79 26

27 BEREC also questions such options in its report: The implications of this measure however go far beyond roaming as any additional competition generated could revolutionise competition in domestic markets [ ]. For these reasons, BEREC is inclined to the view that the measure would be regarded as disproportionate, given that the problem to be solved is high roaming prices 36. BEREC also questions its impact on competition on roaming markets within this report: this would be a very significant extension of the Regulation [ ] it would not itself generate more wholesale competition [ ] 37. Finally, it is worth remembering that if the current regulation on roaming departs from the regulatory framework for electronic communications, it has been so justified by specific reasons (recitals 4 to 13 of the Regulation). This refers to regulation of wholesale markets for international roaming at national level said to be not sufficient, notably because of its crossborder nature, the national regulatory authorities not being able to control the behaviour of the operators of the visited network and roaming markets having specific characteristics justifying exceptional measures going beyond the regulatory framework. These reasons can certainly not be transposed to relations between MNOs and MVNOs operating in the same national market. Those measures therefore raise serious concerns, especially for option (iii), notably in terms of commercial freedom on a competitive market. In any cases, were the Commission interested in supporting options (i) or (ii),, a serious impact assessment shared with w stakeholders would first be required. 36 BEREC aforementioned report on roaming, BEREC aforementioned report on roaming, p

28 III II.. Medium to long term view Wifi networks and all IP data networks Question 13: In the medium to long term, markets and technologies will possibly evolve to the point where roaming services can be provided by different competing technologies. Such developments seem to be unlikely to be sufficient to eliminate or minimize roaming problems within 5 years. Do respondents share this view? Please explain. Question 14: Do respondents think that the Commission should pursue measures to accelerate these developments (e.g. to encourage the massive deployment of interconnected Wifi networks? What other measures could be considered? What will the impact be of the transition to an 'all IP' environment on roaming services? OFTG believes that in the medium or long run roaming services will be provided by various technologies. As shown previously, data consumption is increasing tremendously and will continue to do so for the coming years. The growing usage of data roaming will further increase competition both among operators and technologies. Therefore OFTG fully shares BEREC conclusion, In the longer term, retail data roaming could become a reasonably competitive market, given that there are partial substitutes (e.g. via Wifi access), which provide competitive pressure. Networks are also expected to transition to LTE. If that happens and assuming growth in use of VoIP, outgoing voice could also become competitively provided 38. Taking under consideration growing availability of smartphones, tablets access to data via Wifi becomes very easy. Also VoIP application allowing for video/voice IP based communication are progressively available on different mobile operating systems (ios, Android, Symbian, Bada, etc.) allowing users to talk with each other seamlessly. This will lead to additional competitive access to Internet while abroad. In fact, quality Wifi coverage is already a reality in numerous areas in Europe, and Wifi already constitutes a competitive alternative to data roaming in those areas. OFTG is supporting the development of this technology and is facilitating its use by its customers. For instance, our French customers have access to over 75,000 available partner hotspots worldwide, thanks to roaming agreements. OFTG offers in France more than 30,000 hotspots available to the travellers, thanks to the roaming agreements. Customers are indeed using such opportunities competing with their mobile network, [CONFIDENTIAL]. The real figure is surely much higher as these figures don t include any of the traffic originating from private boxes, free hotspots or competitors hotspots. All this Wifi roaming traffic is data traffic which would have otherwise been sent over GPRS/3G data roaming. Moreover the Wifi roaming services are provided by more and more actors in the public places (i.e. Hotels, airports, railways stations, etc) and private areas like restaurant or coffee chains, which offer free wireless. Free access to the Internet via WLAN becomes a standard option for travellers. Finally the centres of mostly visited touristy destinations are equipped with WLAN connection free to use. OFTG is a supporter of such alternatives: for instance, its Business Everywhere 39 solutions automatically connect by default to the cheapest and best network amongst all those available in the vicinity, regardless of the technology, including GPRS/3G/3G+ services via Orange networks or via roaming partners, Wifi through Orange, public hotspots or private hotspots. 38 Berec report, International Mobile Roaming Regulation p

29 Having in mind that roaming service is inevitably correlated with tourist traffic one can suppose that an alternative connection, such as Wifi access, is easily accessible for travellers. Besides, operators, as well as software developers, are working to ensure seamless transition between roaming/and Wifi based access while roaming, so the customer can save time and money and will not be bothered with linguistic problems. Regarding LTE, the expected rollout should allow this technology to start serving as an alternative to current GPRS/3G technologies within the next 5 years. LTE is already a reality in some countries such as Sweden, Germany and Poland. However, OFTG deems it necessary to point out that an All-IP network such as LTE is different from an All-Internet network, in which the communication would only be done in best-effort, and we are therefore confident that an all-ip network wouldn t lead to a disruption of our current models. The financial and human investments; the issues around spectrum allocation; the inclusion of such services as handover mechanisms that are absent in Wifi and that allow to keep a connection while moving; are many indications that the business models for LTE would be similar to those of GPRS/3G. In conclusion, the data retail market is clearly ly facing an increased competition via Wifi networks and soon via LTE. In order to encourage this technology revolution to happen, the European Commission should ensure a smooth and timely switchover of the analogue broadcasting by

30 IV.. Specific S Issues Inadvertent roaming Question 15: To what extent is the problem of inadvertent roaming still a concern for citizen's living close to borders? What measures could be taken to avoid the adverse effects of inadvertent roaming, whether by means of voluntary co-operation between operators or by means of regulatory or legislative action? Question 16: If you are an operator, what measures (technical or otherwise) have you taken to deal with the issue of inadvertent roaming, both to prevent it happening and to compensate for the adverse effects once it has been shown to have occurred? How do you raise awareness of the problem and the potential remedies on the part of your customers? Throughout the Orange footprint, inadvertent roaming is a very limited phenomenon, occurring at most only a few times per quarter. We believe that sending a welcome SMS to roaming customers proves to be useful as it calls customers attention on the fact that they are on a roaming situation while not being abroad. Apart from this existing measure, and despite Orange s extensive technical investigations, there is no other measure that can help mitigate the risk of inadvertent roaming, as it is mainly a technical issue. Dedicated spectrum engineering has been implemented to reduce this risk. Specifically, our transmitter antennas have been designed in order to comply with standards 41 and to reduce radiation in the direction of the neighbouring countries. However, frequencies will not stop their propagation at borders, and they are not only affected by geographical conditions but also by weather conditions. At different times of day, the strength of a signal can vary significantly depending on the temperature, humidity, etc, rendering such variations unpredictable. In the rare remaining cases of inadvertent roaming, OFTG deals commercially with the impacted customers, as a goodwill gesture, on an individual basis. We understand from BEREC s latest roaming benchmark that this practice is common throughout the industry. OFTG therefore considers c that inadvertent roaming is a mere technical issue that is best resolved individually by mobile operators and does not require any regulatory intervention. 41 including ITU Radio Regulation and CEPT recommendations BEREC Benchmark Data Report for January 2010 June 2010, Oct. 2010, BoR (10) 50, p36,

31 Outermost regions Question 17: What has been the impact on mobile users and service providers of the implementation of the Regulation as far as roaming within, from or between the outermost regions is concerned? Question 18: What additional measures (if any) have been taken by the Member States or their NRAs to address roaming between the outermost regions and other parts of the EU? Additional measures have been taken by the French Parliament concerning the application of the Eurotariff in French outermost regions. Indeed, in 2007, the Parliament approved the extension to French Overseas regions of the prices caps defined in the Regulation. Thus, price caps applicable in the European Union have been extended to French overseas regions without taking into account the additional costs due to the distance from Continental Europe. [CONFIDENTIAL] The Canary Islands enjoy the same conditions: they share the same roaming tariffs as the rest of Spain towards the other European countries, both at retail and wholesale levels, although this territory is also far from mainland Europe. With regard to outermost o regions, OFTG invites the Commission to consider including a specific set of measures, such as a price cap stabilization for these regions.. This would ensure that price caps applicable to outermost regions take into i account their local conditions tions. 31

32 Impact on smaller operators Question 19: What has been the financial impact (revenues, costs, profits, volumes etc.) on smaller mobile telephony providers of the application of the Regulation since its entry into force on 30 June 2007 and amended in 2009? Please provide financial data and any other information in this respect wherever possible (which will be treated as confidential if so requested). Question 20: Has any operator encountered problems when seeking to agree a wholesale roaming agreement with an operator in another Member State? What kind of problems were these (e.g. for SMS interworking)? Were they resolved in the end? Was the issue referred to an NRA? If so, what action has been taken or is in train to address those problems? This question is not relevant for OFTG. However, it worth noting that OFTG did not wait until entry into force of the 2009 Regulation to grant wholesale discounts (including on data) to smaller operators. In effect, in order to negotiate a discount agreement with an operator in a visited country, the main leverage any mobile operator has is his steering ability, far more than its size. A mobile operator s steering ability only derives from its own efforts in deploying the required solution in its network. This then allows the operator to effectively control its outbound roaming traffic, as further described in questions 21 and 22. Regarding Q20, OFTG is not aware of any operators that would have formally complained on wholesale roaming agreement. 32

33 Traffic steering Question 21: To what extent is the use of traffic steering accompanied by a lower retail price for the roaming customer? Where lower roaming prices are conditional upon the use of a preferred visited network, how effective is the traffic steering in practice in ensuring that the preferred network is used? Please provide detailed data where possible. Traffic steering capabilities are to be understood as the ability for an operator to direct outbound roaming traffic based on quality of services and wholesale commercial agreements negotiated. These have enabled operators not only to provide a higher quality of service to customers but also have contributed to providing competitive retail prices for the benefit of customers. It is not OFTG practice and policy to develop network dependent offers for roaming. In other words, OFTG offers are available on any networks throughout Europe and customers will benefit of the same retail prices whatever network they use. We observe that BEREC in its latest roaming benchmark indicated that, Evidence suggests that [traffic steering] is used to provide an enhanced or lower priced service to the roaming consumer 44 and thus that our views are shared by most if not all of the European mobile industry. Question 22: What techniques are applied to implement traffic steering in practice? Is the roaming customer informed in advance about the steering and does he have the possibility to override it? Traffic steering is executed by: Preferred networks list saved on SIM cards A network-based traffic management system Over-The-Air based (remote) updates of preferred networks SIM applets supporting choice of preferred network Roaming customers are not informed about the traffic steering as it has no influence on retail roaming prices. However they are informed that we have preferred partners who offer good or best quality and services. The roaming customer has the ability to choose manually and without fee any specific visited network he prefers, which will override the operator preferred selection. Although preferred networks are in place, this does not prevent the roaming customer from roaming on non preferred networks where the preferred network has no coverage. 44 BEREC Benchmark Data Report for January 2010 June 2010, Oct. 2010, BoR (10) 50, p36,

34 Impact on domestic prices Question 23: Have you identified any significant effects on domestic prices or changes in an operator's tariff structure for domestic voice calls or other mobile services introduced after or shortly before the entry into force of the Regulation? If so, please explain providing details of the changes in terms of timing, scope and prices. In the past, we have not seen significant effects on domestic prices. For the future, in light of some of the EC s options, OFTG stresses that roaming prices shouldn t distort competitive domestic pricing. In that respect, the thorough impact assessment OFTG calls for should take into account the possibility of unintended consequences on domestic retail markets. This is a concern in all cases, whether regulation is a forced convergence of roaming and domestic tariffs, a price cap based regulation, OFTG underlines that domestic retail markets are currently competitive and characterised by offer segmentation (innovation, penetration, coverage, unit prices. ), which meets customer needs but inherently introduces a broad range of price tags for the different components of the offers. Any roaming regulation that would put at risk such segmentation by imposing one type of offer or of price-level would be unacceptable and would go far beyond the objective of roaming regulation, raising questions in terms of proportionality. OFTG is concerned about the possible spill-over of any new roaming regulation to a much more significant part of all domestic markets. In that sense, we fully agree with the criticisms made by BEREC regarding any approach based on prices and conditions similar to those prevailing in the domestic markets. At the very least, it would be a rather bizarre outcome 45 if European consumers [were to] pay less for roaming services than for domestic service 46. In fact, in such a case, roaming regulation would impact domestic competitive markets as domestic prices would tend to align onto roaming prices, by far exceeding the alleged roaming competitive issues at stake. It would thus be in breach of the European framework, notably the principle of proportionality. These criticisms could also be relevant in other regulatory options. Any new roaming regulation should take into account the level of domestic prices and should ensure that roaming regulation would not negatively impact domestic competitive markets and/or place operators under margin squeeze for both domestic and roaming business. OFTG would very much like to participate to any impact assessment the EC will launch aforementioned BEREC report on Roaming, p aforementioned BEREC report on Roaming, p 13 34

35 Impact on international roaming arrangements nts with operators in third countries Question 24: What, if any, has been the impact of the Regulation on reciprocal roaming arrangements between EU/EEA mobile operators and their counterparts in other third countries? Question 25: Have any Community-based providers of mobile roaming services negotiated agreements with third country operators concerning a reduction of wholesale roaming tariffs comparable to those set up in the Regulation? In Europe, an unbalanced situation can occur with some countries not within the scope of the roaming regulation. In some cases, some non-regulated countries (Switzerland) have followed the European trend, being part of OFTG agreements. This also means that Orange s member company in Switzerland offers to almost all Orange s European partners the same commercial terms (well below regulatory caps) as it would in any other Orange regulated country. Conversely, in some other non-regulated countries (Andorra especially), where OFTG is a large purchaser, roaming charges remain extremely high; in this case OFTG (and its customers) is highly penalized. In other parts of the world, the situation also varies. Wholesale roaming agreements have generally followed the same trend as intra EU/EEA roaming agreements, i.e. IOTs have decreased, supported by the extension of steering tools/skills. During negotiations between OFTG and some large countries (often countries purchasers onto Europe, such as USA), intra-european rates have been used as benchmarks and thus IOT decreases have been significant, bringing wholesale discounted rates to levels close to intra-european. The same is probably true for other European operators. However, there are many markets/countries (Maghreb for instance, but also the Middle-East, Russia-Ukraine ) onto which European operators/groups are OFTG is mainly purchaser and where discounted wholesale rates paid by OFTG remain very significantly above the intra- European level. The same is probably true for other European operators. In some cases, OFTG ends up paying wholesale roaming charges higher or equal than /min. Overall, there is a trend whereby intra-europe roaming charges are decreasing constantly and significantly; while roaming charges outside Europe paid by European operators are not following the same trend, leading to a two two-speed speed market. In this context: - A large non-european Group/operator could decide to set up / launch in Europe some kind of MVNO, and penetrate the European market, with the protection of the EU Regulation. But no European Group/operator has a similar opportunity/right in any non- European market (single country or Group of countries, equivalent in scale/business potential). - There is also a risk that retail roaming charges may very strongly differ for the customer, depending on the country where he is roaming: low (or very low) intra-eea/eu and much higher in third-world countries; which could generate a non-eu billshock. - OFTG is a net purchaser 47 onto third-world countries, probably like other large European Groups. OFTG is therefore financially penalized by the asymmetrical financial flows, as wholesale revenues decrease while wholesale costs increase. 47 roaming wholesale charges higher than roaming wholesale revenue 35

36 Appendix 1: OFTG innovative offers the Favo avourite country offers: For a modest monthly fee the customer can make roaming voice calls at a price close to standard domestic per-minute prices. For example, Orange Spain offers roaming calls at 0.21 /min for a monthly fee of 5 /month. In comparison, standard domestic out-of bundle prices of 0.18 /min. Orange France offers to its postpaid customers for the same monthly fee of 5 /month, one hour of roaming incoming calls and roaming outgoing calls at national price (0.37 /min inc. VAT) for one of the following destinations, chosen by the customer: French overseas departments, Portugal, Switzerland, Spain, Belgium, United-Kingdom, Germany or Luxemburg. Orange Austria s Hallo Deutschland offer: For a 25/month fee, Orange Austria offers the Hallo Deutschland offer providing 1000 free minutes and 100 SMS within Austria and Germany and between the two countries. the Destination Europe offer: Similarly to the Favourite country offers, Orange France offers the Destination Europe offer to its postpaid customers. For a 10 /month fee, the customer is offered within Europe one hour of incoming roaming calls and roaming outgoing calls at national price (0.37 /min inc. VAT) Orange Austria s Europe300 bundle: This bundle provides 300 minutes of EU roaming voice traffic (calls made and calls received) in addition to 1000 minutes, 1000 SMS and 50MB of domestic traffic, for a monthly fee of 25. Domestic offers integrating roaming traffic: Orange offers domestic bundles where roaming minutes are deducted from the domestic tariff bundles. Orange France s Origami Star & Destination Europe offers provide for instance Calls from France to abroad at domestic price Calls from abroad at domestic price 1hour per month to receive calls when roaming Origami Star + Destination Europe/ Prefer P referée = roaming at domestic standard rate (0.38 /min) and 1h free to receive calls (like at home) and direct decrement from domestic bundle France Telecom Group restricted

37 Europe ajustable adaptive offer for data roaming for business customers: Launched in February 2011, Orange France s Europe ajustable offers with a 10 monthly fee to business customers doing data roaming in Europe. Prices adapt to the consumer s monthly consumption. Consumption levels vary from 5MB to 3GB, and prices for data roaming traffic drop down to 0.45 /MB whenever the customers generates at least 100MB of data roaming traffic in Europe. An equivalent offer exists for data roaming worldwide. Travel Data Daily by Country Member company Bundle Details Mobistar 2 /2MB Surf & send s abroad while saving up to 72% off the standard rate 2 MB of traffic, sufficient for occasional use and moderate roaming Travel Data Daily is valid for 24 hours Europe, Switzerland, USA and China). No subscription or commitment Benefit from a 50% discount on additional MB over the pass s period of validity Orange France 2 /2MB Pass valid for 24hrs for all your multimedia connections from abroad Valid in over 195 partner networks in 3G + / 3G/EDGE/GPRS 5 /10MB Automatically activated for new residential smartphone subscription Orange Poland Orange Spain Orange Romania Orange Slovakia 8zl /2MB 2 /2MB 2 /2MB 5 /10MB 2 /2MB Renewable 2MB of data to be used for GPRS data transmission in the EU (Zone 1 + Switzerland) Customers automatically receive as standard offer for mail & mobile browsing. No matter the operator to connect you, always pay the same In all countries of the European Union & in Switzerland. Bundle will automatically continue to rollover with a 2 /2MB offer with a validity of 30 days. In all countries of the European Union & in Switzerland. Bundle will automatically continue to rollover with a 2 /2MB offer with a validity of 30 days. In all countries of the European Union & in Switzerland, 2 per month. Above 2MB each additional megabyte of data subscription will cost only 1.99 /MB 37

38 Appendix 2: 2 Ex Market 15: declared competitive by NRA at 1 st countries round of market analysis in almost all 38