Economic Replicability Tests for Next-Generation Access Networks
Social Science Research Network
This paper discusses the relevant cost standard for the economic replicability test for Next-Generation Access (NGA) networks, described in the "Recommendation on consistent nondiscrimination obligations and costing methodologies" adopted by the European Commission. According to the Recommendation, in order to reconcile investment and competition, wholesale prices should have nonlinear characteristics and only partly vary by the number of accesses. It demonstrates that a cost standard for the
... onomic replicability test, which implies fully fixed and variable cost recovery for the access seeker, including the total wholesale price, would be incompatible with the economics of NGA networks and that such a test would deter NGA investment. Therefore, the cost standard should include only the variable part of the wholesale price. However, this paper underlines that during a transition phase, until competitors have secured access to NGA infrastructure, a second temporary test called the "competition migration test" should be added to ensure incumbent NGA retail prices do not foreclose efficient copper-based entrants. The two proposed tests surpass the limits of the "ladder of investment" theory by including the "business migration effect" developed by Bourreau, Cambini, and Dogan (2012) . The Recommendation (see Annex II pp. 27-29 for specifications) suggests that the test should be done with a discounted cash flow (DCF) on an average customer lifetime and account for long-run incremental costs plus (LRIC þ) as a cost standard. At the same time, the costing methodology for the EC should guarantee an "appropriate balance between ensuring efficient entry and sufficient incentives to invest." According to the document, this would imply allowing operators investing in NGA networks a certain degree of pricing flexibility. This flexibility would enable significant market power (SMP) operators and access seekers to "share some of the investment risk by differentiating wholesale access prices according to the access seeker's level of commitment." The EC views volume discounts and/or long-term access pricing agreements as important tools for fostering NGA investment. This implies that the total wholesale price paid by the access seeker is not necessarily strictly proportional to the number of accesses, as it may include elements such as a minimum fee, discounts above certain volumes, upfront payment or co-financing arrangements. This paper models all these forms of nonlinear wholesale access price structures as "two-part tariffs." In this simplified model, the variable part of the wholesale two-part tariff is by definition the part of the wholesale price actually paid by the access seeker that is proportional to its number of accesses. All other components of the wholesale price will be considered as the fixed part of the wholesale price. This type of price structure reflects the underlying investment cost structure for an operator that invests in fibre to replace its copper access network, since a large part of its investment cost is fixed and independent of demand that is otherwise uncertain. In this case, the economic analysis shows that optimal wholesale prices should include a fixed component in order to effectively allocate the risk that fixed costs may not be covered if demand is low. In this context, the proportion of the access network cost that may be legitimately subject to a form of nonlinear wholesale pricing depends of the proportion of the copper access network being replaced with fibre. Thus, when NGAs take the form of "Fibre to the Node" (FTTN) or "Fibre to the Cabinet" (FTTC), for which only a limited proportion of copper is replaced by fibre, wholesale prices are only slightly nonlinear, for instance through limited volume discounts. In this scenario, a large part of the wholesale price stays variable and the fixed part is relatively small. On the other hand, when NGAs take the form of "Fibre to the Home" (FTTH) where the copper access network is completely replaced by fibre, wholesale prices may be much less linear, using approaches like co-financing for the fixed infrastructure, significant upfront payment or more significant volume or duration discounts. In this case, the proportion of the total wholesale price that is actually proportional to the number of access is lower. This paper addresses the question of how to implement the economic replicability test for NGA networks recommended by the European Commission in such a context. The Recommendation itself suggests the use of nonlinear wholesale pricing, but does not consider the implications for the ER test. Determining how to implement the test when wholesale prices are nonlinear requires further investigation. Academics and institutions have analysed "margin squeeze" in depth (for instance, see Jullien, Rey, Saavedra (2013) for an overview), but only for linear wholesale prices. Indeed, the question of how economic replicability or margin squeeze tests should be implemented when wholesale prices are nonlinear has not been formally analysed. Moreover, the recommended test should be structured to fulfil the EC's dual objective of encouraging NGA investment and maintaining the competitive structure inherited from copper unbundling while following the principle of fair investment risk distribution between access provider and access seekers. 1 The novelty of this paper is also the proposal for a solution for two-part wholesale prices, with different and complementary regulatory regimes for the variable and fixed parts of wholesale prices. The paper mainly addresses the case of total replacement of the copper infrastructure by an NGA network. However, the proposal would remain valid in intermediate situations requiring partial replacement of copper networks, but its significance would be reduced in due proportion. Although the paper directly refers to the EC Recommendation on consistent non-discrimination obligations and costing methodologies, the economic arguments would also be suitable for a margin squeeze test under competition law. This paper aims at addressing the key implementation issue of the 2013 EC Recommendation on non-discrimination and costing. Therefore, it does not cover important subjects for access regulation, which are not covered in the Recommendation, such as geographical segmentation. The remainder of the paper is organized as follows: Section 2 defines the notion of "economic replicability test" and details the EC formulation in the Recommendation on Cost Orientation and Non-discrimination. Section 3 is the core of the paper, it demonstrates that a test that implies full fixed and variable cost recovery would deter NGA investment. It shows that to reconcile investment and competition, the economic replicability test should only include the variable part of wholesale prices. However, during a transitional phase until competitors have migrated from copper to NGA networks, a temporary test called the "competition migration test" should be added to ensure that the incumbent does not foreclose efficient copper-based entrants. Finally, Section 4 discusses how the two-test system proposed can be integrated into existing regulatory theory and practice. Economic replicability test: definition and EC formulation of the test for NGA Networks In the Recommendation on consistent non-discrimination obligations and costing methodologies published on September 11, 2013, the European Commission advocates allowing for a certain degree of pricing flexibility for NGA 1 EC Recommendation C(2013) 5761 final, 11.9.2013 on non-discrimination and costing. Annex II "When setting the parameters of the ex-ante economic replicability test, NRAs should ensure that the SMP operator is not put at a disadvantage vis-à-vis access seekers regarding the sharing of the investment risk."