NGA business cases considered

Because of the flexibility of the model and the amount of architectural and regulatory options, virtually a nearly unlimited number of scenarios and solutions can be produced. In order to be able to compare the results of the different approaches not only within one country, but also between the countries, we determined 21 standard cases per country (Table 7), for which we produced the critical market shares (for each of the eight clusters). To keep the architectural approaches (FTTC, PON, P2P) “clean” we did not mix them between the clusters or even within one cluster. We also restricted ourselves to a limited amount of regulatory options, which a priori seemed reasonable to us. Yet, even with these restrictions we produced altogether 126 cases. This number is reduced for those countries where VDSL is considered as being not/hardly feasible for technical reasons (i.e. France and Spain).77 _{The cases can be selected by choosing}

the appropriate combination of input parameters.

The stand-alone operators acting as first movers are the base case of the model. They receive ARPUs as reported by the ECTA members. The only choice to be made is for the basic architecture (3 cases). No regulatory measures are included.

The incumbent as first mover is the first variation of the model which, however, does not require any regulatory options. We take account of the extraordinary income from the sale of a major amount of MDF sites by savings in the investment outlays, which then are distributed over time. Furthermore, we assume the incumbent to have a “competitive advantage” compared to competitors: the incumbent is assumed to be able to use its infrastructure for at least one additional purpose, i.e. the infrastructure can be used for providing (an) additional service(s) or product(s) (e.g. leased lines) which in turn allows a further cost sharing regarding the infrastructure. The ARPU is the same as in the basic stand-alone case.

The second movers are assumed to have 10% less revenue (ARPU) than the first movers. The reason is that it is harder for them to enter the market and they may not be successful by just having different (and better) services. Rather, our observation is that the market entry has to be supported by a price discount.

The VDSL cases for the second mover use sub-loop unbundling (SLU) in all of the cases considered, because the construction of new copper lines would never make sense. Hence all regulatory options considered only cover the backhaul and feeder segments of the subscriber access network.

The PON cases for the second mover allow the choice of using self constructed (or partly rented) or fibre SLU rented solutions. The first variants require a fibre

77 To be more precise, we have reduced the number of cases for France by 8. In Spain we only calculated a single VDSL case, namely the one for the incumbent.

infrastructure to each of the homes covered, the SLU variant allows to rent just those fibre sub-loops which are necessary to connect the own customers. This is comparable to the standard copper SLU case (see VDSL). The regulatory options considered cover all three network segments, but with fibre SLU the solution for the distribution segment is already set.

Within the P2P cases for the second mover the model provides to choose between own (and rented) infrastructure down to the customer homes or rented loops (fibre LLU) at the metro core location. This is comparable to the existing copper LLU at the MDF sites although it may allow more replicability due to the smaller number of metro core locations than MDF sites. The regulatory options considered cover all three network segments, but with fibre LLU the solution for the distribution segment is already set as well.

The case 80% shared infrastructure assumes an efficient regulation where the second movers can rent and share 80% of the infrastructure needed to install own cable from the incumbent or other market players (e.g. sewers owned by the City of Paris). The infrastructure to install cables can be empty ducts, sewers or aerial cabling poles or façade mounted holders, …). Because of the cost differences we use the cheapest solution as long as it is available and add empty ducts up to 80%. Typically the relations differ from cluster to cluster (e.g. the less dense populated the more aerial installations are used). This approach is applied to all relevant network segments. The rest of 20% of the infrastructure has to be constructed by the second mover itself. In the case of VDSL (always with copper SLU) and PON SLU we assume the operators to share the same street cabinet.

The case 20% shared infrastructure assumes a less efficient regulation or less availability of rentable infrastructure, but is comparable to the 80% case in all other dimensions. In the case of VDSL it is assumed that the less efficient regulation does not allow to realize the collocation in a single street cabinet. So a second one has to be installed. Only in the case of PON SLU the collocation in the street cabinet is assumed (you only need the SC with OSDF for realizing the SLU).

The case 80% dark fibre access assumes an efficient regulation where the second mover can rent the amount of dark fibre needed instead of ducts. The amount needed differs between the architectures, since they reduce the fibre demand at the feeder and backhaul segments to a different degree. We only consider cases of dark fibre access in combination with VDSL and PON SLU78_{. The fibre price differs between the clusters}

and considers the different construction approaches and cost of the incumbent (e.g.

78 In the cases of dark fibre access without LLU/SLU it is assumed that all fibres to connect the homes of a cluster are rented or constructed in advance, and anyhow are available independently of the market share achieved. This makes these cases very expensive and we excluded them from the considerations here.

aerial). Once again in the case of VDSL and PON SLU we assume the operators to share the same street cabinet.

The case 20% dark fibre access assumes a less efficient regulation or less available infrastructure, but is comparable to the 80% case in all other dimensions. In the case of VDSL it is assumed that the less efficient regulation does not allow to realize the collocation in a single street cabinet. Thus, a second one has to be installed. Only in the case of PON SLU the collocation in the street cabinet is assumed (you only need the SC with OSDF for realizing the SLU).

The case 80% dark fibre/ shared infrastructure incorporates a combination of duct rental in those segments, where the amount of fibres in parallel is high, and fibre rental, where it is low. We only consider this in combination with SLU cases (PON, VDSL), so the ducts are rented in the backhaul segment (MCL – MDF) and the fibres are rented in the feeder segment between the splitters/ street cabinets and the old MDF locations. This variation may reveal that a differentiated approach can be cheaper than a pure approach (e.g. duct rental only).

Table 7: Main assumptions characterising the different cases of our model VDSL PON P2P VDSL PON P2P 80% shared infrastructure 20% shared infrastructure 80% dark fibre access 20% dark fibre access 80% dark fibre/shared infrastructure 20% dark fibre/shared infrastructure 80% shared infrastructure 20% shared infrastructure 80% dark fibre access + SLU 20% dark fibre access + SLU 80% dark fibre/shared infrastructure + SLU 20% dark fibre/shared infrastructure + SLU 80% shared infrastructure 20% shared infrastructure LLU Remarks Stand Alone operator as first mover 1

(i) 10% less revenues than in (1) and (2)

(ii) In the feeder segment: 80% of deployment is realized by dark fibres (iii) In the backhaul segment: 80% of deployment is realized by empty ducts (iv) 2nd mover and incumbent share the street cabinet

(i) Investment savings due to dismantling of MDFs (ii) Better use of shared infrastructure

(i) 10% less revenues than in (1) and (2)

(ii) 80% of deployment is realized by using air cables and empty ducts (iii) 2nd mover and incumbent share the street cabinet

(i) 10% less revenues than in (1) and (2)

(ii) 20% of deployment is realized by using air cables and empty ducts (iii) 2nd mover uses its own street cabinet

(i) 10% less revenues than in (1) and (2) (ii) 80% of deployment is realized by dark fibres (iii) 2nd mover and incumbent share the street cabinet

(i) 10% less revenues than in (1) and (2)

(ii) 80% of deployment is realized by using air cables and empty ducts

(i) 10% less revenues than in (1) and (2)

(ii) 20% of deployment is realized by using air cables and empty ducts (i) 10% less revenues than in (1) and (2)

(ii) 80% of deployment is realized by dark fibres (iii) 2nd mover and incumbent share the street cabinet (iv) Sub-loop unbundling is considered

(i) 10% less revenues than in (1) and (2) (ii) 20% of deployment is realized by dark fibres (iii) 2nd mover uses its own street cabinet

(i) 10% less revenues than in (1) and (2)

(ii) In the feeder segment: 20% of deployment is realized by dark fibres (iii) In the backhaul segment: 20% of deployment is realized by empty ducts (iv) 2nd mover and incumbent share the street cabinet

(v) Sub-loop unbundling is considered 2nd mover PON

4

2 Incumbent as first mover

(i) 10% less revenues than in (1) and (2) (ii) 20% of deployment is realized by dark fibres (iii) 2nd mover and incumbent share the street cabinet (iv) Sub-loop unbundling is considered

2nd mover VDSL 3

(i) 10% less revenues than in (1) and (2)

(ii) In the feeder segment: 80% of deployment is realized by dark fibres (iii) In the backhaul segment: 80% of deployment is realized by empty ducts (iv) 2nd mover and incumbent share the street cabinet

(v) Sub-loop unbundling is considered (i) 10% less revenues than in (1) and (2)

(ii) In the feeder segment: 20% of deployment is realized by dark fibres (iii) In the backhaul segment: 20% of deployment is realized by empty ducts (iv) 2nd mover uses its own street cabinet

(i) 10% less revenues than in (1) and (2)

(ii) 80% of deployment is realized by using air cables and empty ducts

(i) 10% less revenues than in (1) and (2)

(ii) 20% of deployment is realized by using air cables and empty ducts 2nd mover P2P

5

(i) 10% less revenues than in (1) and (2) (ii) LLU is considered

As sensitivity we add the expenditures of second movers for infrastructure rented from the incumbent to the incumbents revenues per cluster as infrastructure wholesale revenue and investigate the effect on the incumbent’s critical market share and profitability.