1. ITU/BDT Arab Regional Workshop on "Developing Competition Policies and Strategies in Telecommunications" Rabat-Morocco, 19 - 21 December 2005 Mobile market analysis for Ex Ante regulation : The EU case Presented by Dr. Rochdi ZOUAKIA

2. Disclaimer The eventual views expressed are the authors’ own and do not necessarily represent those of the NRA he is representing, its staff, or any of its experts.

3. This presentation analyses the way in which termination rates between any network and a mobile network (the payments made by one operator to terminate a call on another operator’s network) transfer resources between those networks and thus affect output levels, prices, competition and market structures in the telecoms system as a whole. Many studies have addressed the question of mobile termination alone, from the standpoint of identifying the relevant market, identifying the presence of significant market power and, where it is found, considering appropriate remedies. These studies have either taken as given the regulatory package in operation from 1998, or have looked ahead to the new regime in effect within the European Union after July 2003. By contrast, the presentation focus on the simultaneous impact of fixed and mobile interconnection rates in EU countries as they apply to fixed to mobile or mobile to fixed traffic, since these rates can transfer resources between the two sectors and thus affect their relative scale and prosperity.

4. Figure 1. The CPP mobile termination model

5. The regulatory framework has been successful in ensuring interconnection among all networks, fixed and mobile, thus giving consumers the benefit of ‘any-to-any’ connectivity. The very fast growth of mobile connections, now as numerous as fixed lines in many European countries, has been a strong feature of the last ten years. However, the principles governing the regulation of access to fixed and mobile networks have been quite different. In fixed networks, the dominant firm – the historic monopolist has been obliged by the regulator to offer access to its network for the purposes of interconnection or call termination at prices which have usually been set by the regulator to equal cost, including a reasonable return on capital employed: either a direct estimate of network costs or a benchmark based on ‘best practice’ abroad.

6. Other fixed networks typically have to set interconnection charges at the same level as that determined by the regulator for the dominant network. Mobile networks pay the same prices for interconnection to fixed networks as do other fixed networks. Mobile networks, however, have not generally been subject to price control, either of their termination rates or of their outgoing call prices. As a result, they have charged termination rates which exceed estimates of termination costs by a wide margin. Where EU regulators have found a mobile operator to have significant market power, they have had the power to impose cost- based terminations under the 1998 regulatory arrangements, but have not generally done so. In recent years, however, many regulators have put pressure on mobile operators to bring termination rates down gradually, using so-called glide paths, but rates still exceed costs by a large margin.

7. These high, uncontrolled termination rates have applied almost universally to fixed to mobile calls. Practice has differed with respect to termination of calls to mobile from other mobile networks – which now make up 50% or more of all calls to mobile. In some countries, lower termination rates have been charged on such calls than in the case for fixed to mobile calls. In others the same rate is charged on all incoming calls. In one country (France) an originating mobile operator makes no payment to the terminating operators, an arrangement known as ‘bill and keep’.

8. Figure 2. Evolution of fixed and mobile domestic voice traffic in UK

9. One possible justification for high mobile termination rates is that they have generated resources permitting operators to offer handset subsidies and low outgoing call rates, which have brought new subscribers on to the networks and benefited everyone who can call them. But this effect will only operate fully if competition in the outgoing call market is so intense that all such termination surpluses are competed away, which is not generally the case. There is no guarantee that mobile operators will choose the optimal level of subsidy; and the argument for subsidies to bring new people on to the network loses its force as mobile penetration achieves the very high current levels. Moreover, the process of driving down outgoing mobile call rates using termination revenues may have brought fixed and mobile prices into a spurious convergence, encouraging callers to use a mobile line when a fixed line would impose lower costs on the economy.

10. In EU countries, it has been estimated that the scale of the transfer as a result of high mobile termination charges for fixed to mobile calls from fixed networks and their customers has, over the five years 1998- 2002, amounted to 19 billion. The figure is calculated as the excess of termination charges paid over costs, including a normal return on capital employed, in France, Germany and the UK alone. The precise estimate depends on assumptions about costs, but it is clearly substantial.

11. Figure 3. Transfers as a fraction of total mobile revenues (retail + interconnection) in UK, Germany and France

12. The effect of this transfer has been to injure fixed customers and their operators, and it is likely to have damaged competition in the fixed market. The beneficiaries have been mobile operators, which in some cases may have shared some of their gains with their customers in the form of lower outgoing mobile prices. The transfer has also distorted competition between fixed and mobile operators. An optimal policy towards fixed and mobile termination charges will redress the imbalance by setting both on the same basis, eliminating the existing asymmetries between cost-based fixed termination rates and uncontrolled (or loosely controlled) mobile termination rates.

13. However, the new European regulatory arrangements permit a range of possible remedies to be applied once a mobile operator has been found to exercise significant market power in the market for termination on its network. These remedies include setting cost- oriented termination rates, as is the case for fixed networks. Regulators may also want to consider imposing a non-discrimination condition, which requires a network to charge the same rate for termination supplied by a mobile operator to itself (for on-net calls) and that charged to other operators, and to require that a mobile operator’s retail prices for on-net calls at least cover the termination charges it levies on others. Whether concurrent price control is needed on the retail price of calls from the dominant firm’s fixed network will depend upon competitive conditions in retail calls from fixed networks. We do not recommend controlling mobile outgoing prices. These policies will introduce a better balance between the fixed and mobile services, encourage competition in fixed services, promote more efficient pricing throughout the sector, and thus benefit consumers of telecoms services overall.

14. The termination rate issue: The OPTA case After issuing a public consultation an undertaking a comprehensive study regarding mobile termination tariff, OPTA has formulated the policy that constitutes the basis for determining the maximum reasonable MTA tariff, which will be applied as benchmark to assess the reasonability of the MTA tariffs applied by a mobile provider.

15. OPTA came out with a decision regarding Lowering termination rates in mobile networks based on the following: In order to reach a mobile end-user, access must be acquired to the SIM card in the mobile telephone. Only the mobile provider from whom the user acquired the SIM card can supply this access. A mobile provider supplies other mobile providers the opportunity to terminate telephone traffic on its mobile network solely by means of the call termination service. Thus there is no true opportunity for supply or demand substitution with the service call termination. On the one hand, potential competitors cannot offer a comparable service unless they have access to the SIM card. The Commission believes that a substitution of this type is not to be expected because a mobile provider will not be willing to allow competitors direct access to its clients. On the other hand, other providers have no technically and economically feasible alternative service for terminating telephone traffic with mobile end-users. The Commission therefore concludes that in terminating traffic on its own mobile network, each mobile provider holds a (near) monopoly

16. In defining the reasonability with reference to MTA tariffs, the fact that is determinant for the Commission is that mobile providers have no decisive stimuli to lower their tariffs to a cost-efficient level. In the Consultation Document of 19 December 2001, OPTA/IBT/2001/203784, the Commission described this lack of decisive stimuli for more cost efficient tariffs among mobile providers. The responses from the market parties during the consultation gave the Commission no cause to revise this viewpoint. The Commission will describe its considerations with reference to the responses from the market parties in more detail in a separate document. On a wholesale level, the lack of competition in the supply of call termination and the legal obligation to establish interconnection have made it virtually impossible for market parties wanting to terminate traffic. There is an alternative solution, in which telephone traffic from other networks is delivered ‘on net’ to the mobile provider by means of what is known as a SIM switch that is connected to one or more network connection points for end-users.

17. The mobile provider then terminates the traffic as if it were traffic between two end-users connected to its own network. The amount of traffic that is currently terminated using this system is minimal. What is more, the technical and contractual limitations involved make it improbable in the Commission's view that this system will be developed within the foreseeable future into a true alternative for call termination. A network connection point for an end-user, for example, is not suited to terminating a large volume of traffic, and with a view to systems of this type, mobile providers often place contractual restrictions on the issue of SIM cards. At this time, this system therefore does not represent an alternative for call termination. with a mobile end-user to counter the pressure of the market power of mobile providers in terms of call termination. Neither does the behaviour of the end-users stimulate cost efficiency.

18. The mobile provider does not experience any real pressure from the end-users connected to its network. These ‘own’ end-users place insufficient importance on the level of the MTA tariff because they do not have to pay it themselves16. In selecting a certain mobile operator, the MTA tariffs are usually not a decisive factor for this reason. The level of the MTA tariff is often also unclear at the time that the end user selects a mobile operator. A good indication of the lack of influence from the behaviour of the operator’s ‘own’ end-user was seen in actual practice in the Netherlands: when KPN Mobile lowered the MTA tariff and launched an accompanying information campaign, this did not result in a significant increase in demand for call termination on KPN Mobile’s network. The calling behaviour of the end-user calling to the mobile network also offers no real stimulation to determine the MTA tariff at a cost- efficient level. The level of the MTA tariff is often also unclear at the time that the end-user calls a mobile number. This unfamiliarity is stimulated by the lack of transparency often seen in the tariff structure for calls to a mobile number

19. However, the development of competition on the telecommunication market will, in the Commission’s view, not profit from an immediate reduction of the MTA tariffs, which would pose an extreme threat to the mobile providers. In its choice for a differentiated reduction of the MTA tariffs and the introduction of a transition period, the Commission has, in so far as possible, taken into consideration the possible detrimental effects of the reduction in the level of the MTA tariffs on the competition on the mobile market on a shorter term. The Commission expects that reducing the MTA tariff to the level of the mobile providers with the best performance in Europe will bring an average decrease in the net result from mobile call termination of 21%. Due to the decreased income from call termination, the total income of the mobile providers will decrease by 5%27. The Commission believes it is reasonable to give mobile providers the opportunity during a transition period to attune their business plans and activities to these figures. However, the Commission also believes that the interests of end-users and fixed providers must be considered. The market parties will profit from having the tariff reduction implemented as quickly as possible.

20. Taking all aspects into consideration, the Commission has concluded that a transition period of eight months after publication of the policy rules, during which the tariff is reduced in two steps as per 1 May 2002 and 1 December 2002 to the level of the best performing providers in Europe, is reasonable. It must be noted that since publication of the consultation document ‘Regulation of mobile terminating tariffs’ on 19 December 2001, the mobile providers have known the Commission’s intention to reduce the MTA tariffs as well as the nature of the future European ONP rules, which must be implemented by the Member States by mid-2003.

21. Justification of the best practice benchmark applied 1. The Commission has determined the maximum reasonable level of MTA tariffs. This is the level that will be applied as per 1 December 2002 as reasonable MTA tariff; it will be applied in determining the end point of the sliding scale. 2. The Commission relates this level to the tariff level of the best performing mobile providers in European that are not subject to a legal cost orientation obligation. 3. In order to determine this level, the national regulators of the EU Member States as well as Norway and Switzerland were asked to submit the MTA tariffs applicable to mobile network providers. A total of 57 mobile providers are active in these countries. The tariff data from 47 companies, representing 82.5% of all mobile providers active in these countries, were involved in the tariff comparison. This gave the Commission more than sufficient information to arrive at a comparison. At the end of this appendix, the mobile providers included in the study are listed.

22. 4. The data received by the Commission was divided into the tariff elements call setup / conveyance and peak/non-peak. In order to accurately compare the tariffs with those applied by the mobile providers in the Netherlands, it was determined that the data would be averaged based on the traffic pattern of telephone calls that terminate on a Dutch mobile network during the period of one year, i.e. the fourth quarter 2000 through the third quarter 200128. The tariffs of European mobile providers with 900 MHz and possibly an 1800 MHz network were averaged on the basis of the traffic pattern of KPN Mobile and Vodafone, and the tariffs of the mobile providers with solely an 1800 MHz network were averaged on the basis of the traffic pattern of Ben, Dutchtone and Telfort. 5. Traffic pattern is understood to mean here the relationship between traffic in the Peak and in the Non-peak/WNT29 periods. The Commission decided to combine the telephone traffic in the Off-peak and WNT periods because many of the mobile providers in Europe do not distinguish between these ‘off-peak’ periods.

23. 6. For the benchmark, the following calculation was performed to determine the average MTA tariff of foreign mobile providers. First, the turnover for 'the Netherlands’30 was determined for both the Peak and the Off-peak/WNT periods by multiplying the setup tariff by the total number of calls and multiplying the conveyance tariff with the total number of call minutes to mobile networks in the Netherlands. Subsequently, these turnover figures were added and divided by the total number of minutes. 7. This calculation was used to translate the tariff structure of the 47 European mobile network providers to a uniform and comparable average MTA tariff that should be charged in the Netherlands for the ‘average call to a 900/1800 MHz or solely 1800 MHz mobile network’, respectively. These translated tariffs per European provider are hereinafter referred to as the ‘(European) benchmark tariffs’.

24. Figure 4. Example calculation of the average Mobile termination tariffs per minute for a fictitious European provider

25. 8. For calculating the ‘best practice’ level, per country it was determined which provider applies the lowest average MTA tariff. Subsequently, the lowest of these providers in terms of average MTA tariffs were selected. Their tariffs were subsequently averaged to the level of best performing companies in Europe. 9. The mobile providers were divided into three groups: A. providers with no legal cost orientation obligation using a network that is fully based on 1800 MHz; B. providers with no legal cost orientation obligation using a network that is based on 900 MHz or a combination of 900 and 1800 MHz for MTA tariffs; C. providers with a legal cost orientation obligation (in practice, there are no providers with a network that is fully based on 1800 MHz in this group).

26. 10. In the benchmark, only the average MTA tariffs of the providers in the groups A and B were considered. For this purpose, the data from 10 providers in 7 countries in group A and of 23 providers in 11 countries in group B were included in the tariff comparison. 11. The three selected providers and their tariffs are shown per group in the tables below. Based on the benchmark, the Commission arrived at a European best practice level, where an average MTA tariff of € 0.1614 per minute applies for mobile providers in the Netherlands with a network fully based on 1800 MHz, and an average MTA tariff of € 0.1257 for mobile providers in the Netherlands with a network based on a combination of 900 and 1800 MHz. The three selected European mobile providers and their MTA tariffs are shown in the following two tables.

27. Figure 5. European Benchmark best practice 1800 (without cost orientation) Figure 5. European Benchmark best practice 900 (without cost orientation)

28. Conclusion of the OPTA case

29. CONCLUSION Mobile termination rate was in the last five years on of the hottest topics in Europe as far as competition regulation. EU members regulated differently the said termination rate until the general framework has been issued by the commission. By and large, the policy applied by OPTA in the example given above was applied by EU members that is at a certain determined time, termination rates should be cost oriented.