TI Trunk service share calculation method

A5.183 As set out in Section 6, we propose to identify separate TI trunk markets for regional and national trunk segments at all bandwidths. In this subsection, we set out our approach to measuring market shares for these markets.

A5.184 As per our assessment of market shares for terminating segments (AISBO, TISBO and MISBO markets), in our base case, we rely on CPs’ circuit data (retail sales and wholesale purchases and sales) to derive market shares for trunk markets. We have based CPs’ share of circuits in trunk markets on circuit information drawn from exactly the same datasets used to assess shares for other leased lines markets.

Hence, our trunk analysis is based on the data that was obtained, cleaned and checked in line with the steps set out above.

A5.185 The particular circuit information used to assess trunk market shares using this data includes:

• name of the CP providing or purchasing the circuit;

• interface of the circuit used (we assess TI circuits only);

• bandwidth of the circuit;

• location of each end of the circuit (based on easting and northing coordinates and/or postcode data); and

• whether the circuit was sold as part of a VPN-solution.¹²⁷

125 Para 5.32, Consultation on Price Controls for wholesale ISDN30 services, April 2011.

http://stakeholders.ofcom.org.uk/binaries/consultations/isdn30-2011/summary/isdn30-2011.pdf.

126 Para A6.71, Statement on Price Controls for wholesale ISDN30 services,

http://stakeholders.ofcom.org.uk/binaries/consultations/isdn30-price-control/statement/isdn30-annexes.pdf.

127 Under our base case, we exclude circuits used as part of a VPN-solution from our trunk assessment.

However, we have run sensitivity analysis to understand the impact of including TI circuits used as VPN-tails (as discussed below).

A5.186 We have followed five main analytical steps to derive trunk market shares, namely:

i) we identify TI circuits that contain trunk segments;

ii) we determine whether those circuits are used for national or regional trunk;

iii) we apply adjustments to trunk circuit counts to take into account the bandwidth of the trunk circuits sold;

iv) for each CP, we apply equi-proportionate mark-ups to the circuit counts to account for missing geographic and bandwidth information; and

v) we estimate (bandwidth weighted) market shares for the national and regional trunk markets using each CP’s self-supply and its sales of trunk to other CPs to compute its total trunk supply.

We use Trunk Aggregation Node ‘catchment areas’ to identify terminating and trunk segments

A5.187 As explained in Section 6, we identify the boundary between TI trunk and terminating segment markets based on the location of trunk aggregation nodes (TANs).

A5.188 We have identified 46 such TAN locations based on key urban centres where CPs are likely to locate (at least) one of their key interconnect points to pick up

termination traffic. The catchment areas associated with each TI TAN are shown in Figure A5.12 below.

Figure A5.12 TAN catchment areas

A5.189 The catchment areas shown in Figure A5.12 are based on the information BT has provided on its PPC logical routing model, whereby¹²⁸:

• each address in the UK is served by a particular local exchange;

• every local exchange is parented to one of BT’s 67 Tier 1 nodes; and

• every Tier 1 node is assigned to the TAN grouping to which it belongs.

A5.190 So for example, the Southampton/Portsmouth TAN includes all addresses associated with local exchanges logically parented to the Southampton and Portsmouth Tier 1 nodes. In turn, the Southampton and Portsmouth Tier 1 nodes form part of a single Southampton TAN catchment area (TAN number 41 in Figure A5.12 above).

A5.191 On the basis of these TAN catchment areas, then for each CP and for every single circuit (retail and wholesale sales and purchases) we determine whether that circuit contains a trunk segment. We rely on the A-end and B-end address details of each circuit to determine the relevant TAN catchment area. Where both ends of a TI circuit fall entirely within a defined TAN catchment area we count a circuit as a terminating segment only. Consistent with our market definition proposals in Section

128 BT applies this logical routing model for charging purposes to determine the proportion of circuits that contain trunk or terminating segment.

6, we assume that any TI circuit that links different TAN catchment areas contains a trunk segment.

We determine whether TI trunk segments are national or regional

A5.192 Having identified that the TI circuit in question contains a trunk segment, we then identify whether the trunk segment falls either within the national or regional trunk market. Again, we use TAN catchment areas to determine the boundary between national and regional trunk:

• a regional trunk segment is a circuit between adjacent TANs; and

• a national trunk segment is a circuit between non-adjacent TANs.

A5.193 Figure A5.13 shows the 46 TAN catchment areas based on the information BT has provided on its logical parenting of local exchanges back to its Tier 1 nodes. Routes between adjacent TANs, which would be regarded as regional trunk, are shaded differently to routes between non-adjacent TANs which are regarded as national trunk.

Figure A5.13 Adjacent and non-adjacent trunk aggregation nodes

ABERDEEN BIRMINGHAM BISHOPS STORTFORD BRIGHTON BRISTOL CAMBRIDGE CARDIFF/NEWPORT CARLISLE CHELMSFORD COVENTRY OXFORD CRAWLEY CROYDON DONCASTER EDINBURGH GLASGOW/CLYDE VALLEY GLOUCESTER GUILDFORD IPSWICH IRVINE KINGSTON LEEDS LEICESTER LIVERPOOL LUTON LONDON CENTRAL LONDON WEST LONDON EAST LONDON DOCKLANDS LONDON NORTH MANCHESTER MILTON KEYNES NEWCASTLE NORTHAMPTON NOTTINGHAM PRESTON READING SALISBURY SHEFFIELD SLOUGH SOUTHAMPTON/PORTSMOUTH SWINDON WARRINGTON WATFORD WOLVERHAMPTON YORK

ABERDEEN ^blank ^0.01 ^0.01

BIRMINGHAM ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 ^0.01

BISHOPS STORTFORD ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 ^0.01

BRIGHTON ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01

BRISTOL ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01

CAMBRIDGE ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 ^0.01 ^0.01 ^0.01

CARDIFF/NEWPORT ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01

CARLISLE ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 ^0.01

CHELMSFORD ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01

COVENTRY ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01

OXFORD ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 ^0.01 ^0.01

CRAWLEY ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 CROYDON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01

DONCASTER ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01

EDINBURGH ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 GLASGOW/CLYDE VALLEY ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01

GLOUCESTER ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 GUILDFORD ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01

IPSWICH ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank IRVINE ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank

KINGSTON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01

LEEDS ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 ^0.01 LEICESTER ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01

LIVERPOOL ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 LUTON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 LONDON CENTRAL ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01

LONDON WEST ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 LONDON EAST ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01

LONDON DOCKLANDS ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01

LONDON NORTH ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 MANCHESTER ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 MILTON KEYNES ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01

NEWCASTLE ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 NORTHAMPTON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank

NOTTINGHAM ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 PRESTON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 READING ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 ^0.01 ^0.01 SALISBURY ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 ^0.01 SHEFFIELD ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 SLOUGH ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 SOUTHAMPTON/PORTSMOUTH^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank SWINDON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank WARRINGTON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^0.01 WATFORD ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank WOLVERHAMPTON ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank

YORK ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank ^blank

A5.194 Figure A5.13 shows, for example, that we would count circuits between the

Aberdeen and Edinburgh and between the Aberdeen and Glasgow/Clyde TANs as regional trunk, whereas all remaining circuits from Aberdeen to another TAN would be counted within the national trunk market (e.g. Aberdeen to London Docklands).

A5.195 For each CP, we generate two measures of supply:

• a count of TI circuits that contain a trunk segment (i.e. using the first step to identify circuits which run between different catchment areas); and

Regional trunk route

National trunk route

• a count of trunk circuits that belong either in the national or regional trunk

markets (based on the second step that checks whether the circuit ends reside in adjacent or non-adjacent TAN areas).

We apply bandwidth weights to trunk circuit counts

A5.196 As set out in Section 6, in our market definition for regional and national TI trunk services we do not propose to identify breaks in the market by bandwidth. However, if we were simply to count each trunk circuit without adjusting for the bandwidth provided over that link, we may end up with a bias in our market share estimates.

This is because a CP may purchase a 155Mbit/s trunk circuit in the wholesale market and use that trunk segment to deliver a number of lower capacity retail circuits (e.g. a number of 2Mbit/s retail circuits). If we did not adjust for these differences in bandwidth then we would not be assessing circuit sales on a

comparable basis. We therefore adjust our trunk counts by assigning greater weight to higher speed circuits relative to lower speed circuits.

A5.197 The bandwidth weightings we apply use the European Commission’s recommendation on retail leased lines prices as referred to in Ofcom’s disaggregated markets statement¹²⁹ and as used in the 2007/8 Review (see

paragraphs 7.358-7.359). The weightings we use are shown in Figure A5.14 below.

Figure A5.14 Bandwidth weightings applied to trunk circuits (Mbps = Mbit/s)

Source: Ofcom 2006

A5.198 Hence, in our circuit counts, if a CP sold two circuits at 155Mbit/s (which has a weight of 26) and ten trunk circuits at 64kbit/s (which has a weight of 1), we would produce a weighted average supply measure of 62 trunk circuits.¹³⁰

129 See page 98: “Disaggregated markets – leased lines”, Ofcom, Discussion document, March 2006 http://stakeholders.ofcom.org.uk/binaries/consultations/disagg/summary/consultation.pdf

130 2 x 26 (the weighting factor for 155 Mbit/s circuits) plus 10 x 1 (the weighting factor for 64 kbit/s circuits) = 62.

We apply mark-ups to our data to account for circuit entries with incomplete geographic and bandwidth information

A5.199 In our data analysis, we apply equi-proportionate mark-ups to our measures of supply to account for entries in the circuit inventory data with missing, or otherwise incomplete, geographic and bandwidth information.¹³¹

Mark-ups for missing geographic information

A5.200 The steps we follow to calculate whether a circuit contains a national and regional trunk segment rely on geographic information on the A-end and B-end of each circuit (as set out above). However, as discussed from paragraph A5.72 above, in some cases CPs were not able to provide complete geographic information for both ends of every circuit they buy or sell. For circuits with missing geographic data, it is not possible to calculate directly whether or not a particular circuit might contain a trunk segment. But to ensure that our trunk services take into account all circuit sales, and not just those with complete address information, we mark-up CP volumes such that the resulting total supply (of circuits with and without trunk segments) matches the total number of circuit sales in the data inventory.

A5.201 For each CP, we perform separate mark-up calculations in the different circuit categories (retail sales, wholesale purchases and wholesale sales) to account for missing data. This process is exactly analogous to that used in relation to the SBO service share calculations, as described above. We take the total number of circuits with missing data and allocate these in proportion to our estimates of circuits that contain trunk (and in proportion to the split between regional or national trunk segments).¹³²

Bandwidth assumptions for missing bandwidth information

A5.202 As explained above, our market share analysis is based on circuit counts weighted by bandwidth. As such, we also need to apply an appropriate assumption for the bandwidth of circuits with missing bandwidth information.¹³³

A5.203 Where bandwidth information is missing for particular circuit, we have assumed that the circuit sold would be 2Mbit/s. The practical implication of this assumption is that any circuit with missing bandwidth information would be assigned a value of 4 in our

131 The dataset we use to estimate trunk market shares does not require further adjustments to account for missing information on interface types. This is because we allocate all circuits with missing interface types to relevant markets (e.g. to AI and TI markets) at the data processing stage as discussed from paragraph A5.47 above.

132 For example, if a CP sold 10,100 wholesale circuits, and we could identify 1,000 with a national trunk segment, 4,000 with a regional trunk segment and 5,000 circuits with a terminating segment only, then we assume that 10% of the CP’s wholesale sales would be ‘national trunk’; 40% ‘regional trunk’; and the remaining 50% would be terminating segments only. The CP in question had 100 wholesale circuit sales with missing geographic information. We would allocate 10 of these circuits to the national trunk and 40 circuits to the regional trunk market. The CP’s final count of wholesale sales of trunk segments would be 1,010 (national) and 4,040 (regional).

133 The assumption applied to circuits with missing bandwidth information sold is only relevant to the extent that we apply bandwidth weights in our trunk market share analysis. For example, where we do not apply bandwidth weights, we would simply count individual circuits in the relevant trunk market irrespective of the bandwidth of the circuit.

bandwidth weighted circuit counts (consistent with the bandwidth weights set out in Figure A5.14 above).¹³⁴

We estimate wholesale market shares using retail and wholesale circuits A5.204 Analogous to the wholesale SBO markets, we have based our calculation of trunk

market shares on data provided to us regarding CPs’ wholesale and retail activities.

Our estimates of market share are based on the following calculation:

𝑇𝑜𝑡𝑎𝑙 𝑡𝑟𝑢𝑛𝑘 𝑚𝑎𝑟𝑘𝑒𝑡

= 𝐵𝑇 𝑠𝑒𝑙𝑓𝑠𝑢𝑝𝑝𝑙𝑦 + 𝐵𝑇 𝑠𝑎𝑙𝑒𝑠 𝑡𝑜 𝑂𝐶𝑃𝑠 + 𝑂𝐶𝑃 𝑠𝑒𝑙𝑓𝑠𝑢𝑝𝑝𝑙𝑦 + 𝑂𝐶𝑃 𝑠𝑎𝑙𝑒𝑠 𝑡𝑜 𝐶𝑃𝑠 A5.205 In order to calculate market shares, we have therefore required data on sales of

trunk circuits and data on self-supply. The data on ‘BT’s sales to OCPs’ and ‘OCP sales to CPs’ was directly provided to us by CPs. However, we did not obtain direct information on CPs’ self-supply and hence we have had to estimate the amount of self-supplied trunk.

We infer OCP self-supplied trunk from retail requirements

A5.206 We have estimated a particular CP’s self-supply by examining the total trunk

In document Business Connectivity Market Review Review of retail leased lines, wholesale symmetric broadband origination and wholesale trunk segments (Page 70-80)