Response to OPTA’s Consultation on
Collocation and One-off Costs
(OPTA, 2 October 2000)
INTRODUCTORY COMMENTS
WorldCom welcomes the opportunity to respond to OPTA’s consultative document “Collocation and non-recurring costs with regard to access to the local loop”, issued on 2nd October 2000.
We fully support OPTA’s initiative to publish Guidelines for collocation before the end of this year. As OPTA does, we also believe that the non-discriminatory, cost oriented and transparent provision of collocation is a fundamental precondition for the further competitive development of DSL services in Holland.
We also fully support OPTA’s initiative to use this consultation to investigate the so-called “non-recurring costs” and the costs for collocation in more depth. We
seriously believe that KPN’s current charges are not cost-oriented at all and that KPN is extremely overcharging. We therefore look forward for OPTA to formally decide what the real cost-oriented charges should be. We believe this decision on cost oriented charges should be taken before the end of the year.
Before answering the specific questions that are included in OPTA’s consultation document, we would like to make some general observations.
1. Fundamental to competition is that the non-discrimination principle is fully applicable for collocation. Collocation needs to be looked at as a common asset that has to be rationalised between all applicants (including KPN itself) on an equal basis. It therefore must be ensured that KPN treats all applicants (including its own DSL-branch) on a completely equal basis
2. The allocation process should not be approached as an individual but as a collective process, in which KPN as the supplier of collocation should be obliged to meet the demand (of KPN itself and its competitors) for collocation as a whole. We recommend OPTA to take notice of OFTEL’s consultation on the Bow Wave Process1, and install similar proceedings in the Netherlands.
3. The regulatory regime for collocation should incorporate the necessary flexibility to allow the market itself to match demand and supply for collocation. At the same time we also believe - especially at this moment in time with huge collective demand and scarcity-issues to be resolved – that it is crucial that speculation should be avoided. We therefore believe secondary trading should be prohibited, at least temporarily.
4. Collocation involves two key components that both are equally important. The first component is the physical space. The second is information. Both are key products that should be provided by KPN in a transparent and
non-discriminatory way.
5. KPN should be obliged to provide Bitstream Access to other DSL providers for a price based at LRAIC as a precondition to start offering DSL-services to retail customers.
1
6. Besides cost-orientation, non-discrimination is a critical precondition for the pricing of collocation. Effective accounting separation is a key requirement to ensure this.
ACCESS TO THE LOCAL LOOP AND COLLOCATION
QUESTION 1
Means of Access to the Local Loop
WorldCom agrees with OPTA that a distinction can be made between Full
Unbundling of the Local Loop, Shared use of the Copper Line and High Speed Bit Stream Access. We fully support the view that these three services should be made available by KPN and that these services are to be considered as being
complementary rather than mutually exclusive. We agree with OPTA that Bitstream Access does not involve collocation.
We believe that for the further competitive development of the DSL-market in Holland (as elsewhere) it is of crucial importance that OPTA ensures that all operators, including KPN’s own DSL branch, are treated equally.
In this context we would like to stress that KPN should be obliged to provide Bit Stream Access as well as Shared Access whenever and at least as soon as -KPN launches its own DSL retail services in a situation that Full Unbundling is not (yet) available as a credible “make” option for alternative providers. In these circumstances KPN should be obliged to offer Bitstream and Shared Access for a price that is based on LRAIC. Background for this viewpoint is that our experience in the US, Europe and Holland suggests that Full Unbundling presents major implementation challenges even with the benefit of effective regulatory oversight. In particular questions related to collocation and the provision of line testing and management services can give rise to protracted negotiation and delay. Therefore OPTA should make it clear that, whilst each of the three options is complementary to the other, they should also be viewed as evolving one from the other rather than being pursued in parallel along the same time timeline. Bitstream and Shared Access should be a precondition for KPN’s deployment of DSLAM’s and the launch of retail DSL services. In the current Dutch situation this means that KPN should be obliged to offer Bitstream and Shared Access right away.
QUESTIONS 2, 3, 4 Forms of collocation
OPTA makes a distinction between three means of collocation: Physical collocation (including “Porto Cabins”), Adjacent collocation and Virtual collocation. WorldCom agrees with OPTA’s distinction between and description of these three basic forms of collocation. We believe that KPN has the obligation to provide all these different forms of collocation on request.
In our view the general principles for the provision of collocation by KPN are: • Collocation is an essential precondition for ULL
• KPN is obliged to honour any reasonable request for collocation
• When evaluating a request for collocation, KPN should include realistic alternatives for collocation
• Insofar there is shortage of collocation facilities KPN should be obliged to demonstrate such shortage in a transparent and verifiable way. In case of scarcity KPN should also treat itself and other equally.
Notwithstanding the fact that we believe KPN is obliged to offer all means of collocation, until now in most cases we have ordered Physical collocation. Background for this is that from an operational perspective it is essential that the line termination device of the alternative operator is located as close as possible to the Main Distribution Frame (MDF). We strongly believe that KPN should be obliged to offer “Porto Cabins” collocation in case there are no realistic possibilities for (inside) Physical collocation. With respect to “Porto cabins” we feel that OPTA should pay special attention to any possible obstructions by local authorities from a local planning perspective. If the Dutch government is serious on ULL, which we think they are, it is crucial to ensure that there are no restrictions for this specific form of collocation.
WorldCom strongly believes that shared collocation should be offered by KPN, especially in a situation with scarcity.
Another issue we feel OPTA should give attention to is the fact that KPN currently restricts the use of collocation to MDF Access services only2. We believe that this restriction should be abolished. Collocation rights in principle should not be limited to the provision of DSL services only. Operators should also be able to locate other technically compliant equipment in such facilities. In its Guidelines OPTA should clarify that the general principle is that collocation rights are not limited to the provision of DSL services and related equipment to support DSL services. Specifically, alternative operators should have the opportunity to enjoy the same shared costs which KPN derives by locating other equipment – for example modems for the provision of IP services or local voice or leased lines termination apparatus – at the local exchange building. Collocation rights should therefore in principle be “application neutral”. The only constraints should relate to health and safety or compliance considerations and KPN should not be able to dictate the manner in which collocation facilities are utilised by competitors. This viewpoint is also in accordance with OPTA’s own decision of 28 July 2000 on KPN’s RIO, in which OPTA requires KPN to include a generic collocation offer in the RIO.
2
BASIC FACILITIES
QUESTIONS 5, 6, 7
Basic requirements for collocation
We strongly agree with OPTA that KPN should be obliged to define a standard offer for collocation that includes all the minimum requirements. The definition of such a standard offer should be in line with the draft Regulation of the European
Commission.3 We therefore believe that the definition of such a standard offer should be part of OPTA’s Guidelines as well and should be available by 31 December 2000 at the latest.
Please find below a list of further specifications that we believe should be part of KPN’s collocation offer.
1. Dual a/c power supply and distribution boards, with 24 hour emergency generator back up.
2. HVAC capable of cooling the 10 square meters to a temperature of 22+/- 2 Centigrade at an Rh of 40 – 60% with N+1 chillers, based on a heat load of 350 watts per square metre.
3. Ingress protection to IP65, and underfloor water leak detection system.
4. A VESDA fire detection system set at its most sensitive levels, with a normal fire detection system, linked to the local fire station.
5. A 1 hour fire integrity for the collocate rooms, from slab to slab, including under the raised floor. A minimum slab to slab height of 3.5 metres.
6. General floor loading of 5.5kN/m2, with a maximum rack load of 12kN/m2. 7. A telecomm earth of 1 ohm or less.
8. A false floor of 600mm with anti-static tiles and earthing of pedestals every square metre.
9. A small, simple BMS system, monitoring the environment, dc plant , ac supplies etc, signalling back to the WorldCom 24 hour manned control point.
3
The Minimum list in Draft Regulation on Unbundled Access to the Local Loop of the EU for collocation services includes:
- Information on the notified operator’s relevant sites
- Collocation options at the sites indicated under point B5 (including physical collocation and, as appropriate, distant collocation and virtual collocation)
- Equipment characteristics: restrictions, if any, on equipment that can be collocated; - Access conditions for staff of competitive operators;
- Safety Standards;
- Rules for the allocation of space where collocation space is limited
10. Door access control and monitoring.
11. Extinguishers, battery backed main lights as well as emergency exit lamps, emergency call point and escape instructions in dual languages (German, English).
12. Normal lighting, 30% emergency lighting for 30 minutes and an escape route diagram in the room. If the remainder of the building is not lit to the same levels, then an emergency, mains charged handlamp should be provided in the collocate room.
13. The equipment should be powered by 48 volt dc power plant with 4 hour battery back up.
14. Fibre entry to the site and room shall be by 2 separate routes. 15. A wall-mounted telephone should be located near the door.
We feel KPN’s collocation offer should also include a Service Level Agreement (SLA) that should also address ongoing maintenance of the collocation. We also feel the delivery process can and should be improved substantially and that the roll out schedule of KPN with regard to collocation should be more transparent.
Of course it needs no say that an offer for collocation in a room with several inches of water on the floor – which is the case at a location in Amsterdam - is something that is not acceptable at all.
QUESTIONS 8, 9 Collocation sharing
WorldCom supports OPTA’s opinion that it is up to the collocating operator to determine how he wishes to make use of the basic facilities. We also support the view that if an operator opts for an enclosed (caged) space for his equipment, this operator should pay for the supplementary costs thereof, since only the cageless variant should be included in the standard offer we discussed above.
We also fully support OPTA’s preliminary viewpoint that it is up to the collocating operators to decide whether they want to share certain facilities with other
operators, and that this sharing may concern basic as well as extra facilities (space, racks, transmission, etc). In principle the collocating operators should have
complete freedom to decide which facilities will be shared. This means that KPN should not be able to restrict shared collocation by any means and that the current contractual limitations should be eliminated.4 Alternative operators should have the possibility to order collocation by sharing a cage or otherwise enclosed collocation space with one or more other collocating parties pursuant to arrangements between that operator and such third parties, including sublease agreements.
4
In this context we would like to point out that we strongly feel that speculation of collocation space should be discouraged in an effective way. In our opinion this means that secondary trading of collocation space should initially (as long as there is scarcity) be forbidden. We could imagine this restriction to be lifted after an evaluation of what happened in the first year. We also feel that for reasons of efficient use of scarce resources the instalment of a “use it or loose it” rules is legitimated. But at the same time we feel that it is important for operators to have enough flexibility to match supply and demand by themselves, which means that operators (serious operators, not speculators) amongst themselves should be able to re-allocate space. To ensure that these kinds of transfers do not involve
REASONS FOR REFUSAL OF ACCESS
QUESTION 10
Reservation of space by KPN
We fully agree with OPTA that a request for collocation is a reasonable request for special network access. Given the characteristics of the services to be provided on the basis of access to the local loop there is no alternative to collocation available, except a complete duplication of the local loop, which is economically not feasible. In principle we also agree with OPTA that refusal of collocation space can only be based on grounds of scarcity. But this raises the question in what specific
circumstance scarcity is a reasonable argument for KPN to reject requests of alternative operators for collocation. As everybody understands scarcity can also artificially be created by KPN and willingly used as an anti-competitive instrument. This phenomenon needs to be worked out in more detail in OPTA’s Guidelines. WorldCom’s view is that KPN should be obliged to offer collocation to alternative operators and to itself under completely equivalent terms and conditions, and that if there is scarcity it must be secured that the consequences of this collocation-scarcity are also equivalent for KPN Retail and its competitors.
For “new” collocation space this means that everybody, including KPN itself, has to bid for collocation space (including reservations) under equal conditions. In principle this means there should not be any possibility for KPN to reserve any space in advance. Further we feel it is crucial that one should pay for reservations. Herewith one stimulates the efficient behaviour with regard to reservations. In the case of KPN as a vertically integrated dominant enterprise, these costs for reservations should only be recovered via its DSL-customers only. We further believe these costs for reservations should also be included in any future OPTA price squeeze test for KPN’s DSL-pricing).
Regarding “existing” collocation space this means that if alternative operators are (temporarily) hindered in their rollout of DSL-services, due to collocation-scarcity, this should also have an impact on KPN. As OPTA also mentions the situation at this moment in time is that KPN itself is already offering DSL-services under the name MX-Stream at locations, where other operators can not start offering these services due to collocation scarcity. WorldCom believes that this is a clear example of how KPN is (ab)using its position as a vertically integrated company. We
therefore strongly request OPTA to pay substantially more attention to this issue in its Guidelines and develop practical and transparent rules to prevent such abusive “first mover advantages” of KPN. In our opinion a minimum requirement is that KPN should be obliged to offer Bitstream Access to its competitors as a precondition for its own DSL-service launch. Furthermore we believe that OPTA should focus its investigations for spare capacity on these locations from which KPN itself already started offering DSL-services to end-users, and that the results of these
Further we believe that KPN should be obliged to explain the reasons for any reservations in a transparent and verifiable way. Only on the basis of that it can then be decided if the amount of space and the time frame of the reservations is reasonable in relation to the amount of space reserved for third party collocation.
QUESTION 11
Inefficient use of collocation
We fully agree with OPTA’s viewpoint that inefficient use of collocation by an operator that hinders the service-provision of others, should be prevented in an effective way, as long as there is scarcity. As we have argued above under questions 8 and 9, we strongly believe that up-front rules to prevent speculation should be installed, meaning secondary trading of collocation space should be prohibited, at least on a temporary basis. We would also support the instalment “use it or loose it “ restrictions.
QUESTIONS 12, 13, 14, 15, 16, 17 Space for Physical Collocation
WorldCom agrees with OPTA that KPN, in the case of an investigation into the possibilities of physical collocation, is bound to take into consideration all available spaces in an exchange building, and to thereby examine whether (parts of) these spaces are suitable, or can be made suitable, for physical collocation.
We also agree with OPTA’s viewpoint that in principle all the space in KPN’s
buildings, including space and rooms where only KPN itself is collocated, should be made available for Physical collocation of third parties. It is fundamental to ensure that KPN is treated equally as its competitors.
Further we also agree with OPTA that KPN Telecom must also take into
consideration spaces in which equipment is still situated though no longer in use, as well as spaces which are used inefficiently and where extra space could be
realised.
In principle we also agree with OPTA’s suggestion to install an obligation for
clearance of space in case of inefficient use. We however believe that the obligation for KPN itself could be tougher than for alternative operators. As we have already mentioned in the above under 11, we believe that some sort of “use it or loose it” rule can be installed for collocation parties. We doubt a deadline of 3 months is reasonable and feel a deadline of 6 months for collocating parties might be more realistic, taking into account the operation aspects regarding the installation of equipment.
We do not agree with OPTA that if there were no realistic possibility for Physical collocation, (only) virtual collocation would be a reasonable alternative. In sum, our preference for the different forms of collocation is: Physical (inside) collocation, Physical (outside – “Porto cabin”) collocation, Adjacent collocation, (Bitstream Access), Virtual collocation.
QUESTIONS 18, 19
Space for adjacent collocation
WorldCom agrees with OPTA that it is not realistic at all to expect that there will be no available space in an exchange for a connection point for the purpose of
adjacent collocation. However, we do not believe that Porto cabins can be viewed as an alternative option for adjacent collocation. As we have expressed in the above, we believe Porto cabins should primarily be considered as an alternative for (inside) physical collocation.
QUESTIONS 20, 21 Space on the MDF
WorldCom is worried by KPN’s claims that the lack of space on the main distribution frame at a large number of locations may cause bottlenecks in the development of local loop access-based forms of service provision by other operators. In our opinion this claim is unjustified, since KPN itself is the one that causes this potential problem. On the one side, the potential lack of space on the MDF is mainly due to the inefficient way KPN Retail currently uses the MDF for its line-sharing service “MX-Stream”. On the other side, the provision of a line-sharing service by KPN would largely solve this potential lack of space. We therefore do not believe that in the current situation it is necessary to invest in an expansion of the MDF.
QUESTION 22 Other facilities
(Business Confidential WorldCom)
TRANSPARENCY AND NON-DISCRIMINATION
QUESTION 23 Non-discrimination
WorldCom believes it is fundamental for the further competitive and
non-discriminatory development of the market that KPN Retail is treated in exactly the same way as its competitors are treated; not only regarding the availability of space, but also regarding information, pricing and all other conditions. WorldCom fully agrees with OPTA’s viewpoint that the non-discrimination principle means that equivalent collocation service-provision is the only effective way to ensure non-discriminatory treatment of alternative operators and KPN. We also believe that agreeing on this principle is relatively easy, but that more important is how this principle is applied in practice.
If we look at the practice (the market situation) right now, no other conclusion can be drawn than that there is no equivalence in the way KPN treats its own affiliates and alternative operators. There is no equivalence regarding information,
rooms/space (KPN is collocated in different rooms), regarding procedures (KPN itself does not have to order collocation in the same way as we do) and regarding the timing of collocation (it seemed like KPN was already collocated in many of the local switches, before other could order). The result of this discriminatory treatment is that KPN Retail at the moment is the best positioned DSL-provider in Holland capable of rolling out its DSL-services as a the first mover in the market. WorldCom believes this situation is clearly in violation the non-discrimination principle and we strongly believe this should be corrected by OPTA by whatever means necessary. The correction should address the “past” (meaning the sides, where KPN is already collocated) as well as the “future” (new sides, where KPN is not collocated yet). For “new” collocation sides we believe it is fundamental that everybody, including KPN itself, it treated equally on all relevant aspects (information, ordering,
procedure, pricing, space, timing, etc). This means that we expect KPN to order and secure collocation facilities on exactly the same terms and conditions as other operators and to participate fully in the procedures to determine the phasing of KPN’s build and the allocation within sides. As we have already expressed in the above this also means that there should not be any possibility for KPN to reserve any space in advance.
For “existing” collocation sides this means that if alternative operators are
should focus its investigations for spare capacity on the from which KPN itself already started offering DSL-services to end-users, and that the results of these investigations should be disclosed to other applicants for collocation in these sides. We also believe that KPN should be prohibited to do any other allocations in these specific sides as long as scarcity exists and that a strict “use it or loose it” rule should apply for any KPN reservations in place.
QUESTIONS 24, 25, 26, 27 Initial Information
We agree with OPTA that KPN should be obliged to provide information. As we have pointed out in the above the provision of information is another key product that should be provided by KPN in a transparent and non-discriminatory way. Information about the MDF sites, coverage, maps of postcodes covered by particular MDF’s, etc is a crucial precondition to enable alternative operators to evaluate the market-potential for MDF-Access prior to investing in equipment and resources. Furthermore it is of major importance this information is correct, especially because this information if the key-input for business-plans of market-parties.
We believe that the information should include: • Postcode coverage or MDF to postcode mapping
• Location and postcode associated with the MDF’s and the name of the associated local switch
• Size of the MDF’s, ie. Number of lines • Maps
• List of Street Names
• Number ranges associated with the MDF’s • Information concerning Line characteristics • Line length distribution per MDF
• Type of disturbers per MDF
• Distribution of each class of spectrum management per MDF
• In order to cope with the evolution of the local loop architecture it is also necessary to receive information regarding proposed future changes of the network.
We fully agree with OPTA that together with the information described above KPN should also provide information with regard to the availability and/or feasibility of collocation facilities in the different exchange buildings.
It is fundamental that KPN as the supplier of collocation meets the aggregated
demand for collocation. OPTA’s suggestion to oblige KPN to carry out a study on
the availability and/or feasibility of basic facilities in all its exchange buildings now and independent of any individual requests is therefore fully supported by
operators.
QUESTIONS 28, 29, 30
Feasibility Study for extra facilities
We agree with OPTA that if an alternative operator considers ordering additional facilities, it should be possible to submit a request to KPN Telecom for a feasibility study at any time and that KPN should react within one month at the least. Such a feasibility study is strictly required because the alternative operator needs to know what is possible and on what costs, before deciding to order these extra facilities. We also agree that if KPN comes to the conclusion there is no possibility for providing the requested facilities, KPN should be obliged to state its reasons in a transparent and verifiable manner.
QUESTIONS 31, 32 Verification
We agree with OPTA that it is crucial for alternative providers to be able to verify both the initial information and the results of KPN’s feasibility study. A “walk-through” is a crucial requirement to enable this. We believe KPN should be obliged to provide a walk-through upon an individual request within 5 working days.
Furthermore we fully support the instalment of a “trusted third party” for mediation in case of disputes between (individual) alternative operators and KPN. Essential requirements for such a “trusted third party” are:
• Independence; • Expertise;
• Fast mediation process (within 1 month)
The instalment of such a trusted third party leaves apart that a complaint to the (OPTA) should always be possible.
QUESTIONS 34, 35
Order and delivery process
We agree with OPTA that a transparent, simple and flexible ordering procedure must be included in KPN’s reference offer. We believe the process from request for collocation space to installation of equipment should contain at least the following steps:
• Feasibility request
• Reply to request, including results • Order for collocation space
• Preparatory work, ready for installation of equipment
• Delivery of access, upon completion of preparatory work, for installation of equipment
DETERMINATION OF TARIFFS AND COST ALLOCATION
PRINCIPLES
QUESTION 36
Application of the non-discrimination principle
WorldCom believes that it is of utmost importance that OPTA ensures there is competitive neutrality regarding the way KPN treats its own retail arm and its competitors. In our opinion KPN itself should be obliged to purchase collocation at equal costs and under equal conditions as other alternative operators. We therefore fully agree with OPTA that there is no justification whatsoever for the fact that KPN’s retail arm might be paying substantial lower costs for collocation. This is clearly in violation of the non-discrimination principle.
We fully agree with OPTA that this discriminatory situation must be corrected. We therefore supports OPTA’s proposal to oblige KPN to pass on the cost advantages KPN Retail enjoys as a consequence of the special position it has to the alternative operators who purchase collocation. Only by averaging the costs related to
collocation for KPN Retail with the cost related to collocation for other parties the existing situation can adequately be corrected. In practice this will probably mean that alternative operators will enjoy a certain discount that is related to the cost-advantages KPN Retail has enjoyed until now. We suggest OPTA (as OFTEL has already decided) that non-discrimination should over-rule cost-orientation in the sense that the costs charged should be the lower of the cost-oriented rate and the non-discriminatory rate (i.e. the rate that KPN charges itself).
Furthermore we fully agree with OPTA that the costs which KPN must charge to KPN Retail (including the same rate of return as charged to others) must be passed on to the end users of KPN Retail’s DSL services, and not to the end users of the public telephone services of KPN. Otherwise KPN would be in a position to cross-subsidise its DSL services from its public telephone services, which would be abusive. Appropriate accounting separation rules should be put in place to prevent this. This is also of utmost importance with regard to the danger of predatory pricing and/or price squeezing.
We suggest OPTA (as OFTEL has already decided) that non-discrimination should over-rule cost-orientation in the sense that the costs charged should be the lower of the cost-oriented rate and the non-discriminatory rate (i.e. the rate that KPN
charges itself).
QUESTION 37
Costs for basic facilities
While in principle an efficient market would allow KPN to recover the (forward looking) cost of its investment plus an appropriate share of common costs and a reasonable rate of return, we believe such concepts are redundant when
The cost of space rental, the provision of basic facilities and the costs of additional requirements can all be benchmarked against commercial property values and the rates charged by contractors. These competitive market rates would implicitly include a reasonable rate of return on assets.
The issue of common costs arises only insofar as certain facilities (such as lighting, heating, air conditioning and basic amenities) are shared between different
operators (including KPN). In this case, OPTA has two options. The first is to charge each operator for an appropriate proportion of these shared costs (we will deal with considerations relating to the appropriate sharing mechanisms below). In this case, the relevant cost for KPN’s investment in such facilities would be the forward looking cost. The second possibility would be to use commercial property rental rates for improved facilities generally expressed as a cost per sq. metre. Again, the primary considerations are cost-orientation and non-discrimination; if KPN charges itself less than the cost-oriented rates, this should be reflected as a downward adjustment in the rates charged to other operators.
QUESTION 38
The level of cost determination
This question relates to the issue of averaged as opposed to bespoke pricing. Where a decision is made to charge on a bespoke basis, the question arises as to whether prices should vary on a case by case basis, or whether they should be de-averaged according to some other criterion, such as geography. The following broad considerations should be borne in mind when choosing between bespoke and averaged prices:
• Bespoke pricing results in greater scope for abusive pricing, for example
through allowing the dominant operator to raise its rivals costs, owing to its lack of transparency;
• It is more difficult to ensure that bespoke costs are non-discriminatory, in the sense that it is more difficult to use KPN’s internal transfer charge as a benchmark;
• Bespoke prices result in less certainty for competitors to KPN, an important consideration in a market that is dominated by a single operator.
On the other hand, where there are large differences in costs, there may be grounds for allowing bespoke pricing.
pricing should only apply (if at all) to services over and above the basic set of services.
Having concluded that such charges should be averaged rather than bespoke, a second question arises if the charges are above KPN’s internal transfer charge. As stated above, we believe the charges for collocation should be the lower of the cost-oriented rates (as determined by externally benchmarked commercial rates) and KPN’s internal transfer charge.
QUESTIONS 39, 40 Determining tariffs
This section considers whether (and where) prices should be charged on a recurring basis or a non-recurring basis. This is a complex question and one that must be answered in the light of a finer breakdown in the types of costs involved. For these purposes, we believe it is important to distinguish between:
• Per exchange costs (the cost of configuring a given exchange for the provision of collocation. This will include the cost of building a general collocation area suitable for providing the basic set of collocation facilities)
• Per operator costs (the incremental costs of providing an additional operator with collocation, once an exchange has already been set up to provide general collocation)
• System set-up costs (the costs of devising a nation-wide system for ordering and provisioning unbundled local loops, the most significant component of which will be the OSS)
As we have argued above, we believe that per exchange costs should be provided on an averaged rather than a bespoke basis (subject to robustly quantifiable
geographic differences in cost). The question remains as to how such costs will be recovered from operators entering markets (and hence utilising such services) at different times.
Incremental per-operator costs are likely to be small in relation to the other two, (although if additional services are purchased such as a discrete collocation room, they will be larger) and would mostly consist of one-off costs associated with providing the collocating operator with the appropriate electrical connections. Such costs could be recoverable through non-recurring charges.
Regarding per-exchange costs, the issues are considerably more complex. The key questions relate to whether costs should be recurring or non-recurring, and how they should be allocated to operators entering the exchanges at different times. The options would appear to be:
etc. The recurring charges already incurred by the first operator would not be recovered from subsequent operators;
• The recurring and non-recurring per-exchange costs are recovered through a recurring charge to all operators. The charge is calculated by dividing the per-exchange costs by a forecast of the number of operators availing themselves of collocation. Errors in the forecast can be corrected through, say, an annual adjustment in the charges for the following year in respect of forecasting errors made in the previous year. Two variations on this theme would appear to be available. Firstly, the forecasts could be exchange specific, or they could be national (or for sub-national regions).
The first of these options would involve placing all the risk on collocating operators and for this reason it cannot be recommended. The second, which has been implemented in the United States would allow for a far more equitable distribution of risk and therefore far more conducive to a competitive market.
This question is also strongly linked to the question of the costs of vacancy, on which we comment below.
We would argue that system set-up costs should be recovered from all operators (including KPN) on two grounds. The first is that the benefits of local loop
unbundling are seen not only by the customers of competitors to KPN, but also by the customers of KPN itself. It would therefore seem unreasonable to require competitors of KPN to bear the full cost of system set-up, as this would present KPN with an unfair competitive advantage in the market. The second is that most of the system set-up activity will be undertaken by KPN. Unless KPN bears some of the cost itself, it will have no incentive to minimise such costs.
From the consultation document we learn that OPTA seems to be in favour of a non-recurring charge, instead of a recurring charge for collocation. We do not agree with OPTA’s preliminary view. We believe the approach should be based on a more principle approach for the recovery of costs associated with the provision of
collocation. In Attachment I you find a paper of WorldCom in which the principles for recovering collocation costs are discussed in more detail. We believe that the risks associated with collocation are similar to property development insofar as KPN incurs costs in the provisioning of facilities, which it then must seek to recover from those using the facilities. Our position is that KPN should bear the risks associated to collocation and that the current non-recurring KPN charge for collocation should be should be replaced by a recurring charge per square metre. We also believe that the physical space costs should never be higher than the costs of industrial property space in similar locations. OPTA’s observation that a non-recurring charge does not seem to be an obstacle for the further development of DSL-services in our opinion is of no use, since at the moment there simply is no other alternative in place which means that alternative operator are simply forced to “accept” KPN non-recurring charges.
up general collocation areas for physical collocation. Creating such areas will reduce both the time interval and the costs of hand-over of collocation. We believe the costs for vacancy should be recoverable through a national
recurring charge per square metre. We strongly reject OPTA’s idea of passing on costs on a one-time basis per location to the first collocating provider (including KPN Telecom), with pro-rated compensation from the next collocating parties. This system passes on too much risk on the first collocating party. We feel KPN, as the provider of collocation space, should bear this risk. For further arguments please see Attachment I.
As we have explained in the above (our answers to question 10) we strongly believe that reservation of space should also be charged. Introduction of a charge for reservation will stimulate efficiency.
QUESTIONS 43, 44
Cost allocation principles
As we have pointed out in the above we believe these costs should be converted into a national recurring charge per square metre.
QUESTIONS 45, 46, 47 Non-recurring costs
We strongly support OPTA’s proposal to include the issue of non-recurring costs in this procedure and hope and expect that this will enable OPTA to take a decision on a very short notice. The current non-recurring charges of KPN (in NLG and excluding VAT) are given in the Table below.
TABLE – KPN’s current Non-Recurring charges (in NLG, excluding VAT)
NON RECURRING CHARGES PROPOSED BY KPN NLG, excl VAT Line charges
Access Line connection 295,=
Access Line disconnection 295,=
ICO cancellation 180,=
Tie Cable (TC) Provision charges
Initial TC for Physical Collocation (400”) 23.000,=
Additional TC for Physical collocation (600”) 29.000,=
Discount per 100” cable in case of Partially fulfilled order for Physical collocation
3.000,=
Initial TC for Adjacent Collocation (400”) 51.000,=
Additional TC for Adjacent Collocation (600”) 61.500,=
Discount per 100” cable in case of Adjacent Fulfilled order for Physical Collocation
5.250,=
CDF Charges
CDF Provision 7.850,=
Testing
Miscellaneous
Incorrect Fault Test Report 110,=
WorldCom’s opinion is that KPN is significantly overpricing its non-recurring charges. The Table below shows that of the operators listed, KPN’s connection charges are higher than all operators except BT, whilst disconnection charges are the highest.
TABLE – Line transfer costs in different countries
Line Transfer BT5 KPN HKT DTAG Telecom Italia US ILEC average Connection cost (per connection) 493 295 155 239 364 83 Disconnection cost (per line) 76 295 129 132 34 QUESTION 48
Costs for extra facilities
We can not support OPTA’s view that it would be impractical for KPN to deal with a large number of sub-contractors. This assertion seems to have been made on the assumption that shared collocation will not be allowed and that therefore a larger number of sub-contractors will be necessary. We believe that if shared collocation is allowed this will not be the case, and a much smaller and easily managed number of sub-contractors will be necessary. We believe, therefore, that the
operators should be allowed to propose own sub-contractors, or be fully involved in a tendering process, to ensure that a competitive environment is maintained. KPN should not be allowed as a monopoly supplier of building work, with no incentive to act efficiently. The advantages outlined above in our view out-weight the possible disadvantages of security and integrity.
If OPTA would decide otherwise, which we would very much regret, it is crucial that is ensured that KPN acts efficiently, especially since the costs have to be paid by the collocating party. If KPN is the monopoly supplier, it is of utmost importance that KPN provides transparent information on all relevant issues so the collocating party can understand the costs and is able to inspect all relevant documentation, such as tenders and/or cost specifications. Therefore, KPN should be obliged to provide such information and if there is a dispute on the price of the additional facilities, the requesting party must have the opportunity to have a countercheck done by a third party, after which the fairness can be determined.
WorldCom believes the rental charge should be based on the lower of KPN’s internal transfer charge (which may or may not be equal to book value) [as OFTEL decided in the case of BT] and market value, since this is the only way to prevent KPN from pricing discriminatorily. The possible disadvantages OPTA foresees (inefficient use and stockpiling), can in our opinion easily be addressed by the instalment of “use it or loose it” and the prohibition of secondary trading. We therefore do not agree with OPTA’s implicit suggestion that market value would be more effective regardless of KPN’s internal transfer charge. Any “market valuation” of collocation space should be based on rental prices of comparable objects, such as industrial property (office space is not the right benchmark). Furthermore, the correct market benchmarks should be used: commercial property rates are quoted for improved and unimproved space, the former including basic amenities such as walls, partitions, air conditions, basic utilities and conveniences. Where these are charged for separately, the appropriate rates would be those of unimproved space. The benchmarks quoted below are for improved space.
TABLE – Industrial Property costs per square metre per annum (in NLG)
Location Costs per m2 (in NLG)
London West End 510
Amsterdam 193
Hong Kong 246
Frankfurt 407
Milan 136
Although OPTA only requests us to respond to the periodical rental charges of KPN we take the freedom to give some additional comments on collocation charges KPN has put in place. KPN’s current (contractual) charges for collocation are given in the Table below. The prices are in NLG and excluding VAT.
TABLE - KPN’s current charges for Physical and Adjacent Collocation (in NLG, excluding VAT) Non Recurring Charge Yearly Recurring Charge Yearly Recurring Charge per Ampere PHYSICAL COLLOCATION Standard Charges
Facility (standard configuration) 200.000,= Facility (half of standard
configuration)
120.000,=
Cabinet Set (per cabinet footprint) 9.450,=
Optional charges
48v DC no break installation 30.600,=
48v DC no break usage 300,=
230v monitored supply installation 17.500,=
ADJACENT COLLOCATION Standard charges
Facility k 1200 22.500,=
Facility k 2400 23.600,=
Rental charge for k 1200 3.100,=
Rental charge per k 2400 3.250,=
ADDITIONAL CHARGES
Service facility postcode Coverage 100,=
If we compare the above charges to the KPN charges for Interconnection
collocation there is only one conclusion possible; KPN is extremely overcharging collocation for MDF-Access. Current contractual charges for interconnection collocation are given in the table below. In Attachment II KPN’s invoice to WorldCom for Interconnection collocation is presented. .
TABLE – KPN’s charges for Interconnection Collocation (in NLG, excluding VAT) Non Recurring Charge Yearly Recurring Charge Yearly Recurring Charge per Ampere PHYSICAL COLLOCATION
Facility (standard configuration) 0
Footprint per m2 including power 2.425,=
WorldCom has also made a comparison between the collocation prices in Holland, Germany and the United Kingdom. We have compared the following four categories of prices: (1) Exchange related Recurring Costs, (2) Line Related Recurring Costs, (3) Exchange related Non-Recurring Costs, and (4) Line Related Non-Recurring Costs. Furthermore the costs for each of these categories are given in (a) US $ Per Line, (b) Per Exchange, and (c) Per Square Metre. A summary of the outcome is presented in the Tables below. Further details are provided in ATTACHMENT III.
TABLE - Exchange Related Recurring Costs (US Dollars per annum) Per Line Per exchange Per M2
BT 5.3 1,582 329.6
KPN 51.9 15,568 3,243
DTAG 4.78 1,435 299
TABLE - Exchange Related Non-Recurring Costs (US Dollars per annum) Per Line Per exchange Per M2
DTAG 398 119,257 6,625
TABLE - Line Related Recurring Costs (US Dollars per annum) Per Line
BT 175
KPN 185
TABLE - Line Related Non-Recurring Costs (US Dollars per annum) Per Line
BT 437
KPN 341
Regarding copper pair costs, WorldCom commissioned Analysys to conduct a study of country specific metallic path costs for France, Germany, Italy, Netherlands and the UK. Analysys's model of the economic cost of copper local loops was developed in June 2000. It is a parametric model, designed to calculate the costs of copper local loops and to assist those trying to understand the structure of these costs. The model was presented at a seminar in Brussels in July 2000, at which OPTA staff was present. The results are presented in the table below. The further details of the Analysys model and study are already in OPTA’s possession.
TABLE – Costs for Copper Pair (Annual Rental per pair in US Dollars)
Actual or PTT Proposed Analysys
BT Proposed 149.8 101.6
DTAG (Actual) 135.8 150.5
FT (No FT Proposal) N/A 118.2
KPN Proposal 185.06 117.2
Telecom Italia Proposal 138.6 102.6
Notes on these figures. The Analysys model results do not include common costs between the metallic path facility (MPF) and other network services (such as those shared with the MPF (particularly local exchange site costs). In most cases, the rates proposed by PTTs will include the common costs. Therefore, an adjustment has been made to the Analysys figures to estimate the cost including common costs. To make the adjustment, the Analysys numbers were divided by 0.87. The basis for this number was OFTEL’s document “Access to bandwidth: indicative prices and pricing principles” in which OFTEL stated that common costs accounted for 13% of the price BT proposed for the rental of the MPF.
ATTACHMENT IV includes a calculation of what would be reasonable room build costs. This attachment is derived from earlier comments that WorldCom and other UK-based operators send to OFTEL in response to OFTEL’s consultation Access to
6
Bandwidth: indicative prices and pricing principles. Please note that the calculations are in UK Pounds; we believe the UK Pound exchange rate is currently overvalued at around 30%.
The main conclusion in the light of the above are:
• The non-recurring line-related costs for KPN are high relative to other operators we have considered, especially disconnection costs;
• The implied cost of space rental proposed by KPN (NLG 9,450 per cabinet footprint) is extremely high. A typical cabinet has a footprint of 1.2 square metre. The implied rental is far in excess of that charged by itself for other forms of collocation, of any other operator and commercial property benchmarks; • The price proposed by KPN for the MPF is far in excess of our best estimate of
the cost of providing this facility, provided by Analysys. In terms of its impact on total costs, this is by far the most important component of all PTT provided facilities.
QUESTION 52
Other components of Monthly rental charge
WorldCom believes that the monthly rental charge should only include costs related to the actual renting of the space plus the directly related generic costs. Costs that are caused by individual parties should not be included. In our opinion costs of cleaning and security should be included in this rental charge under the
precondition that the costs for these services are reasonable. We would like to underline that costs related to the maintenance of the building should be excluded. These costs should be paid by KPN since the cost of maintaining the fabric of the building in our opinion is a sunk cost, which KPN would incur regardless.
QUESTION 53 Costs for electricity
WorldCom fully supports OPTA’s viewpoint that KPN should charge collocating operators the actual price KPN has to pay for electricity. However, we would like to stress this should be the actual price including discounts KPN might enjoy. We assume that in many cases KPN will have negotiated discounts due to the level and timing of peak demand and that it is possible that KPN will receive extra discounts due to the extra demand generated by the operators equipment. We believe that supplying power at the actual price paid by KPN (including discounts) will ensure competitive neutrality. KPN should not be allowed to extract monopoly rents acting as a reseller of an essential input.
QUESTION 54
Costs of Information - WorldCom agrees with OPTA that there should not be charges related to the initial information. KPN should provide this information for free to the alternative providers.
Cost for the Initial Investigation – WorldCom agrees with OPTA’s preliminary conclusion that there should not be a specific charge related to the initial
investigation. Indeed the information resulting from this study, independent of the question of which party initiates this initial study, is of interest to KPN itself as well other parties which might want to collocate at a certain exchange building. We agree with OPTA that KPN should handle these costs as part of the realisation of the basic facilities in the particular exchange building.
Costs for Feasibility – WorldCom agrees with OPTA’s suggestion for the costs for the feasibility study. In the light of the principle of cost causation it seems
reasonable to decide that if parties come to the conclusion, based on this study and/or a possible ‘walk-through’ procedure following the study, that there is no potential for collocation, the requesting party should pay the costs and if the conclusion is there is potential for collocation the costs should be for KPN.
ATTACHMENT I
PRINCIPLES IN RECOVERING CO-LOCATION COSTS Note by MCI WorldCom International
MCI WorldCom (MCIW) is preparing papers on a variety of pricing issues which arise in connection with the implementation of Unbundled Local Loops. This paper considers the principles which should inform the recovery of non-recurring (and, to a lesser extent) recurring costs associated with the preparation and provision of co-location facilities. It does not concern itself with the pricing methodology (LRAIC vs ECPR), which is the subject of a separate paper, nor does it seek to derive prices themselves (which is the subject of a separate study which MCIW has
commissioned from independent consultants, the results of which are expected to be published in June).
Cost recovery and risk
The issues concerning the recovery of co-location will be familiar to anyone involved in property development. High fixed costs are incurred at the inception of the project to design, prepare and commission the building facilities for which tenants are then found. Costs must be incurred and work undertaken before future revenues can be assured. Space can be left idle as developers mis-forecast demand or prospective tenants change their minds.
Co-location activities are similar to property development insofar as the PTT incurs costs in the provisioning of facilities which it must then seek to recover from those using the facilities. It also incurs costs in conditions of uncertainty since it cannot (and should not) make the availability of co-location facilities conditional upon there being adequate demand to fill all available space thereby provided.
Both property developers and PTTs share the characteristic of incurring both non-recurring costs associated with the design and fit out of facilities and non-recurring costs associated with maintenance, light and air conditioning etc.
However, PTTs differ from property developers insofar as they are being obliged to provide co-location facilities and are otherwise unlikely to do so on a voluntary basis. Any risks are thereby imposed upon them, rather than assumed on a voluntary basis. However, PTTs also differ from property developers in being dominant in the provision of these facilities and it is this dominance which provides the rationale for regulatory intervention, both in terms of setting (or adjudicating upon) the level of charges for such facilities and in terms of allocating the burden of risk and uncertainty.
Models for risk allocation
benefits of competition resulting from Unbundling, principally in the form of lower costs and product innovation amongst all participants in the market including the PTT, are enjoyed by the PTT’s customers as well as those of the third parties. These benefits might, however, be recognised by requiring the PTT to assume the risks of co-location, without their necessarily assuming the costs.
1. PTT offloads risk
PTTs could assume no risks by requiring the first requester of co-location facilities to meet the full costs of the preparation of such facilities. Whether the first requester subsequently sought to recover some of those costs from other third parties is likely to be a matter between new entrants and one on which European regulators will be reluctant to intervene. The extent and manner in which they might expect to be able to recover such costs will determine the risks assumed by the first requester. IN effect, the PTT is off-loading the full cost of the risk to the first requester in these circumstances. The PTT will then be indifferent as to whether that requester has persuaded other parties to underwrite or assume some of that risk.
This model allows the PTT to fully recover their costs in a risk free manner. As a result it has several deficiencies. The first is that it negates the first mover advantages which early adopters of unbundled loops might otherwise expect to enjoy and which would otherwise encourage rapid adoption of Unbundling. A cost recovery model which discourages early adoption of new technologies – and penalises those that do by requiring them to assume high risks – is not consistent with the objectives for unbundling which regulators hold.
In addition to discouraging early adoption, the model discourages widespread adoption. The first requester may seek to recover costs from subsequent users – but even if they incentives to encourage adoption they themselves have no control over the facilities concerned (which remain in the hands of the PTT). Conversely, the PTT has every competitive incentive to minimise the number of users of such facilities – once they have fully recovered their costs from the first requester. Again, this model produces perverse incentives by discouraging (or at least not encouraging) the PTT to fill whatever co-location facilities are provided.
The model could further encourage the inefficient provision of co-location facilities by encouraging over-supply of facilities in the first place – since the higher the initial costs incurred by the PTT the greater the entry barrier to any requesting party under this model. Much depends in this instance upon whether the co-location facilities are dimensioned by the PTT without regard to forecasts from third parties. If the PTT is required to build only in accordance with forecasts then this model could in fact result in under-provision as requesters seek to minimise their risks or to develop them in a modular fashion (by placing small initial orders and then requesting additional facilities – requiring additional development – at a later date). Whether under – or over-supply of co-location facilities is to be preferred will depend upon relative costs and benefits.
2. PTT assumes risk
the opposite – the PTT now has incentives to fill space and to fill it quickly, thereby delivering the competition benefits which Unbundling is intended to secure.
The disadvantages of this approach are that the PTT may tend to under-supply co-location facilities – and to develop in a much more cautious fashion than would be the case if the risk were offloaded to third parties at the outset. This could result in delays or shortages of facilities whilst PTTs seek to only to expand when demand is readily apparent.
The question as to how costs are recovered as between different residents is potentially complex – but no more so than in the case of conventional property development in which rents are assessed by reference to floor space or other considerations. These challenges are also present when the PTT offloads the risk to the first requester – except that in this case the requesting parties then have to agree these principles amongst themselves rather than with the PTT.
ATTACHMENT II
KPN’s charges for Interconnection Collocation (Business Confidential WorldCom)
ATTACHMENT III
Detailed calculations of ULL benchmarks, BT, KPN, Deutsche Telecom
Explanatory note: the INPUT, is in italics and underlined, all other figures are calculated
assuming 1.2m2 per cabinet footprint, 4 cabinets per exchange, 4.8m2 required per exchange, and 300 lines per exchange
Note that the space taken up through each cabinet footprint assumes shared collocation space, figures would be different for
dedicated collocation rooms
BT room build costs - a range of figures were given: worst case is assumed.
BT US$
Per line Per exchange
Per m2 Exchange Related Recurring costs
Collocation space rental per year 4.8 1,438 299.7
Maintenance per year 0.5 144 30.0
Total exchange related
recurring costs (not including power)
5.3 1,582 329.6
Line Related Recurring costs
Copper Rental 172
Tie Cable Rental 3
Total Line Related Recurring costs 175 Exchange Related non-recurring
Room build 437 130,956 27,282
Line Related Non-Recurring costs
Connection (line transfer without NTE) per line £
195
Tie Cable Setup costs per line £ 28 Total Line Related Non-Recurring
costs
KPN US$
Per line Per exchange
Per m2 Per cabinet footprint Exchange Related Recurring costs
Charge for cabinet footprint 51.9 15,568 3,243 3,892
Total exchange related
recurring costs (not including power)
51.9 15,568 3,243 3,892
Line Related Recurring costs
MDF Access basic Rental Dfl per line per annum
185.3 Total Line Related Recurring costs 185 Exchange Related non-recurring
Collocation setup 275
82,371 17,161 AC Power Installation 24 7,207 1,502 DC Power Installation 42 12,603 2,626
Total exchange related non-recurring costs
341 102,181 21,288
Line Related Non-Recurring costs
Connection Charge 121
Internal Tie cable connection charge 24 Total Line Related Non-Recurring
costs
145
DTAG US$
Per line Per exchange
Per m2 Exchange Related Recurring costs
Collocation space rental per year 4.72
1,417 295
Maintenance per year 0.06
19
4 Total exchange related
recurring costs (not including power)
4.78 1,435 299
Line Related Recurring costs data needed Exchange Related non-recurring
Room build: 398 119,257 6,625
ATTACHMENT IV
Room build costs
We believe that the room build costs described in the OFTEL document are very high and imply a large degree of over engineering. Using the example of a standard collocation room of 12 square metres, we believe that an appropriate range of costs would be £26,500 to £31,500. The BT cost range in the OFTEL document is
£31,400 to £76,000.
We have made this assertion on the basis of a comparison of BT’s estimate of room build costs with an estimate of room build costs made by Kingston Communications plc. The estimate is based on similar activities carried out in the Kingston upon Hull network, a smaller network than BT’s. Furthermore, BT must be assumed as an efficient operator to be able to achieve greater economies of scale.
The details of this estimate are illustrated below:
Item BT Cost Range (£k)
Kingston Cost (£k)
Comment
BT Planning activity 2 – 5 0.9 Note 1
Agents fees (for building work assessment)
Π2 - 5
3m x 4m x 4m partitioning (4 walls) + flooring (BT allow £200-£500/sq m)
2.4 – 6 5 Note 2
Electrical fit-out Π2.4
Internal door 2 – 5 1.5 Note 3
New External door to building with break through
2 – 5 3.6 Note 4
Break through from MDF with fire stopping
• 1 - 3 Note 5
Cable runway (100m) 10 - 25 2 Note 6
Access control (x 2) 5 - 10 2 Note 7
Miscellaneous 8 - 20 3
Contract supervision Π3.1
Total 31.4 - 76 26.5 - 31.5
Œ - BT item costs assumed to be included under Miscellaneous • - BT item cost assumed to be covered within Cable runway Comments
1. Kingston allowance is 2 days (16 hrs at £35 per hour + £100 travelling)
2. Use of false floors would be avoided wherever possible in view of timescale and cost
5. Higher cost applies if breaking through equipment strength floors/load bearing walls rather than non load bearing walls
6. BT's cable runway costs seem quite excessive. Kingston's are based upon: • £25 per 3m length of runway
• 2 ceiling sockets & hangers every 2m @ £5 ea. • fixings at 6 per hr, @ £35 per hr including overheads 7. Access control assumes (per door):
• Shear locks at £300 • Card reader £200 • Incidentals £100
• Fitting and testing takes 8 hours @ £35 per hour
We note that BT's proposal to charge distant located operators for a cable runway is unreasonable as there is no need for such a cable runway.
