National Leased Lines   

A Further Statement Issued by the Director General of Telecommunications

November 1999

Chapter 1 Introduction

Chapter 2 National Leased Lines – International price comparisons

Chapter 3 State of competition in the provision of leased lines in the UK

Chapter 4 Conclusion – Review of Regulatory Framework

Oftel is launching a review of the regulatory framework relating to national leased lines. The plan is to complete the review and publish a Statement by October 2000.

It is Oftel’s goal to achieve the best possible deal for UK telecommunications customers in terms of quality, choice and value for money. Oftel recognises that leased lines are an important service which many UK businesses purchase, and are a particularly important cost input for the development of e-commerce in the UK. The Review will consider what are the appropriate changes, both regulatory and de-regulatory, to the current regulatory framework in order to deliver the best possible deal to customers.

This review is a major new project for Oftel. It will take forward the competition concerns and the scope for de-regulation outlined in this Statement; all issues related to the pricing of national leased lines raised in the context of the Price Control Review; and those raised by the Commission’s recent Recommendation on leased line interconnect pricing.

The project will carry out a detailed investigation into BT’s costs and prices for private circuits. It is anticipated that Oftel will work with consultants to develop a suitable model to analyse prices and costs for private circuits. Discussions with operators will be held to assist Oftel in understanding plans for infrastructure development and requirements in relation to a wholesale interconnect leased line product for the provision of end to end leased lines and in relation to the provision other services such as Frame Relay, ATM and internet access. The project will also consider to what extent it is appropriate to de-regulate some markets, for example in terms of price regulation and price publication.

The need for the Review has arisen as a result of market and technical developments since 1996, when Oftel’s current regulatory framework was established, which means that Oftel is now concerned that its current regulatory framework may no longer be appropriate.

Oftel’s review carried out in 1999 of international price comparisons suggests that, while the UK is broadly comparable with other European countries, its competitive position has been eroded and, in relation to the US, it appears that for most types of user US leased line price levels are below those in the UK.

Oftel has found that there are competition concerns relating to the terminating segments (connections from a customer’s premises to an operator’s trunk network). In this market BT has, and seems likely to retain, market power. This suggests a need for targeted regulation.

By contrast, in the market for trunk segments (connections between trunk exchanges) the 1999 review found competition is developing and there do not appear to be competition concerns. This suggests that there may be scope for de-regulation.

There are also other developments which suggest the need for the new Review. Oftel is aware that the state of competition in the provision of leased lines has an important bearing on the state of competition in other markets, such as the provision of Frame Relay, ATM services and internet access, for which leased lines are an important input.

Also, responses to Price Control Review: Future Developments in the Competitiveness of UK Telecoms Markets, published in July 1999, which was the first stage of consultation in Oftel’s review of retail and network price controls, revealed concerns by users and operators about the pricing of leased lines. In particular, issues about the competitiveness and pricing of the terminating segments of private circuits and the relative price levels of short and long distance circuits were raised.

The new Review will deal with all these issues.

1.1 Oftel is launching a review of its regulatory framework relating to national leased lines. This will take forward the competition concerns and the scope for de-regulation discussed in this Statement, all issues related to the pricing of national leased lines raised in the context of the Price Control Review and those raised by the Commission’s recent Recommendation on leased line interconnect pricing. The scope and timescale of the Review are discussed further in Chapter 4.

1.2 The need for the Review has primarily arisen as a result of the further work which Oftel has carried out investigating its concerns and those expressed by users and operators since it published a summary of its investigation into national leased lines in the UK in January 1999 (‘the previous Statement’).

1.3 Oftel has reviewed further international price comparisons and the state of competition in the provision of leased lines. Chapters 2 and 3 of this Statement summarise Oftel’s findings in these two areas.

1.4 The other developments relevant to the Review are discussed in Chapter 4.

1.5 The current regulatory framework specifically relating to national leased lines consists of a retail price cap of RPI + 0% for inland leased lines with a capacity of 64kbit/s or under and no obligation on BT to offer a wholesale leased service to other operators at interconnect prices. This framework was put in place following Oftel’s review of regulatory controls on BT’s prices in 1996 and so reflected Oftel’s assessment of the prospects for infrastructure competition, competition in provision of leased line services and the state of technology at that time.

1.6 There are also general regulatory rules relevant to BT’s pricing of leased lines, including the provisions in its licence which prohibit undue discrimination between customers and require price publication (including any discounts offered).

1.7 The Leased Lines Directive (Directive number 92/44/EC amended by 97/51/EC) also places obligations on operators deemed by the national regulatory authority (NRA) to have significant market power (SMP) in the supply of leased lines in a particular geographic area. Oftel has determined for the purposes of this Directive that BT in respect of the UK (excluding Hull) and Kingston Communications in respect of the Hull area have significant market power. BT and Kingston are therefore obliged to provide a minimum set of leased lines consisting of ordinary quality voice bandwidth analogue, special quality voice bandwidth analogue, 64kbit/s digital and 2Mbit/s digital presentations. The Directive also requires that the prices for these circuits are transparent, non-discriminatory and cost-oriented.

1.8 BT’s rental private circuit prices vary according to distance and bandwidth. In addition, BT currently, operates two pricing zones in relation to its rental charges for inland private circuits, Central London Zone (‘CLZ’) and non-CLZ. CLZ encompasses all the exchanges within the 0171 area. Circuits fall within the CLZ zone where both ends of the circuits are within the 0171 area. Within each zone BT’s prices are geographically averaged for any given length and bandwidth. BT has introduced a change in its pricing structure for private circuits of 34mbit/s and above which will become effective on 1 December 1999. Prices are still nationally averaged for a circuit of a given length and a specified zone within which the customer’s premises reside.

2.1 The most contentious and important area in any discussion of the competitiveness of leased lines relates to price. In particular, are the prices of leased lines in the UK higher than the equivalent price in other countries? This question is in fact extremely difficult to answer. This is confirmed by the numerous consultancy reports and comparisons which exist, many of which appear to give contradictory information as to how prices in various countries compare.

2.2 There are a number of reasons why it is so difficult to make international comparisons of telecoms prices. It is important that these reasons are understood before interpreting the comparisons. This chapter therefore begins by setting out the main problems associated with constructing price comparisons before explaining what the latest pricing data shows. The full structure of this chapter is as follows:

• difficulties of making price comparisons;

• price comparisons within the European Union; and

• UK and US price comparisons.

2.3 Each circuit is priced individually with the price varying according to location, distance and capacity. This means that it is necessary to base comparisons on certain pre-defined criteria (in terms of distance and capacity) in order to ensure that the analysis is comparing like with like in each country.

2.4 Although defining these criteria is straightforward in itself it is important that a representative mix is chosen which covers the full range of circuits available from each operator. If this is not done and, for example, only one or two types of circuit are considered, a misleading conclusion might be drawn. This is because the different pricing structures mean that a number of operators are likely to be able to select a specific circuit for which they are particularly cheap. Unbalanced tariffs, whereby short distance circuits are subsidised by longer distance circuits or vice a versa, make international comparisons particularly difficult. Therefore it is not possible to draw any general conclusions from direct comparisons of circuits of a particular length.

2.5 A second, more significant problem concerning pricing structures is the existence of discount schemes. These are extremely prevalent in the provision of leased lines and are often based on the volume of the order and/or length of contract. The problem with incorporating these discount schemes is that there can be significant variations between operators in how they are applied. This means that it becomes much more difficult to make comparisons on a consistent basis.

2.6 Many of the price comparisons currently in the public domain get around this problem by excluding discount schemes altogether. Although it can be argued that this approach means that each operator is being treated consistently, it is not a particularly suitable solution since it also means the comparisons are not based on what users actually pay.

2.7 This problem becomes even more significant for operators that offer unpublished discounts and tailor packages specifically for individual users. Such bespoke deals are very common in the US and can result in some very large discounts for particular users. However, by definition, such discount schemes are applied on a case by case basis and it is therefore impossible to incorporate them into generic price comparisons. This greatly devalues the usefulness of any comparisons as it further increases the difference between what the data show and what consumers actually pay.

2.8 Published comparisons tend to be based on the incumbent operator in each country only. Such comparisons are appropriate when considering routes on which only the incumbent operates. However, new operators are increasing market share across many OECD countries and in most cases are undercutting the incumbent. The prices offered by these new operators should therefore be taken into account.

2.9 Notwithstanding these issues, Oftel constructed some international price comparisons for inclusion in this Statement. These are outlined in the remainder of this chapter. Much of the analysis is taken from consultancy reports produced by Tarifica, Eurodata and OVUM. In addition, Oftel commissioned Tarifica to look in detail at comparisons between the US and the UK in order to try to better understand how the various discount schemes are applied to different types of user.

2.10 Comparisons of leased line prices in European countries tend to be more straightforward than comparisons involving the rest of the world. This is because the discount schemes offered by European operators tend to be fairly similar in structure and bespoke deals are not too widely available for most types of circuit. Having said this, there are undoubtedly some inconsistencies within the data and the following comparisons therefore need to be interpreted with a degree of caution.

2.11 Figure 1 shows the list price of a 64kbit/s leased line at 50 km and 300 km as at 1 January 1999. It shows that the UK is in line with most other European countries, being very slightly cheaper than the average for both types of circuit. It is worth noting, however, that BT prices have changed only very slightly between 1995 and 1999. This contrasts with an average fall of over 40% for both types of circuits across the EU as a whole over this period. Indeed BT prices for 64 kbit/s national private circuits are basically unchanged since December 1991, although this does of course mean that prices have fallen in real terms.

Source: The Tarifica Leased Line Report 1999

2.12 The above comparisons do not take into account any discounts that are offered by the operators and are thus likely to overestimate what customers actually pay.

Source: The Tarifica Leased Line Report 1999

2.13 Figure 2 gives list prices of a 2Mbit/s circuit across the EU as at January 1999. It clearly shows that for short distance circuits (50km) the UK is one of the cheapest countries within the EU and well below the EU average. This contrasts with the position for long distance circuits (300 km) where BT is above the EU average. This reflects the different pricing structure that BT has adopted compared to some of the other operators.

2.14 The difference in pricing structures is also visible when looking at very short circuits. Comparisons of 2km 2 Mbit/s circuits carried out by Eurodata indicate that prices in the UK are well below those in most other European countries.

2.15 Although BT prices for 2 Mbit/s circuits have fallen slightly since 1995, prices have fallen much more rapidly in the rest of Europe. This is not surprising given the rapid liberalisation of telecoms across Europe over the last couple of years which resulted in some sudden sharp price falls. Within the UK, although the overall price of 2Mbit/s leased lines has fallen, the price of local ends has actually increased slightly over the last couple of years as rebalancing has taken place.

2.16 Technological advances are also driving price falls in many countries. In particular, new entrants are using modern technology to lay huge amounts of trunk capacity on high density routes. As a result these new entrants have a much lower cost base than existing operators and are thus able to offer much lower prices for main links. These costs are becoming less distance dependent and instead leased line prices tend to be driven by the traffic volumes on each route and the price of the local end. It seems likely that eventually price comparisons based on distance, such as are shown in the above charts, will become redundant as prices converge for all lengths of circuit.

2.17 It is difficult to construct reliable international price comparisons for circuits of 34Mbit/s and above. This is because a number of European operators choose to price such high capacity circuits on an individual basis and as a result there are often no list prices on which the comparisons can be based.

2.18 Further, there is a question as to the validity of such price comparisons when based only on incumbent operators. This is because high capacity circuits are, in most countries, one of the most competitive areas of the market which means that the incumbent’s market share can be lower than for other types of circuit. As a result it is more important that international price comparisons of 34Mbit/s circuits and above take account of the prices offered by new entrants to the market. Given these difficulties Oftel has not presented any price comparisons of high capacity circuits in this Statement.

2.19 The above price comparisons are mainly based on analysis carried out by Tarifica. In addition Oftel subscribes to price comparisons of leased lines produced by Eurodata and OVUM. Although these other sources use slightly different assumptions and methodologies they both show a reasonably similar pattern of results to the Tarifica analysis. This leads Oftel to conclude that, bar some fairly minor differences in pricing structure, prices for leased lines in the UK are broadly in line with comparable countries in the rest of Europe.

2.20 This is a change in the position from a few years ago when BT prices were quite significantly below the prices charged by most of their European competitors. The recent liberalisation of telecoms across Europe has resulted in some large and rapid falls in prices which means that BT can no longer claim to be one of the cheapest operators in Europe but is rather in line with the average.

2.21 Users have drawn particular attention to price comparisons with the US. Such comparisons are clearly important. The US is undoubtedly the driving force behind the massive expansion in data communications and in particular the internet and e-commerce for which leased lines are a key input. It is therefore important to try to understand how the US and UK compare.

2.22 Price comparison with the US is extremely difficult. This is confirmed by the studies that exist which often appear to give completely contradictory information about the position of the US in price comparisons of leased lines. For example, according to OVUM (which bases comparisons on prices paid by large business users) the US is more expensive than the UK, Germany and France for most types of national circuits. This is very different to the Eurodata and Tarifica comparisons which generally find the US to be cheaper (and in some cases much cheaper).

2.23 There appear to be four main reasons for these inconsistencies:

• the range of different operators in the US;

• the choice of route being compared (in terms of density of traffic);

• discount schemes; and

• individual deals for different types of user.

2.24 In order to try to obtain a fuller understanding of comparisons between the US and the UK, Oftel commissioned a study from Tarifica that would focus on specific routes in the US and UK and identify how much different types of user would actually pay on each route. The intention therefore was to take full account of discount schemes and the deals obtained by different types of customers.

2.25 It has proved extremely difficult to get the information Oftel required. In particular, it was not possible to establish how unpublished discounts are applied to different types of customer as US operators tended to be unwilling to disclose this information. Indeed, some operators would not provide any information to Tarifica other than to say that all leased line prices are worked out through individual negotiation with the client.

2.26 Tarifica collected what usable tariff data they could. In doing so it became very apparent that prices paid by consumers varied considerably according to a variety of different factors which were unrelated to any quoted list price. For example, one sales employee said that if another operator offered a customer a lower price, he would go to his manager and see about arranging an undercutting price usually by about 10%.

2.27 The most important determinant of the price of leased lines in the US appears to be the level of competition on a particular route. As one would expect high density routes (such as Denver-New York) tend to be cheaper than routes where there is not much traffic. Distance is also a factor in determining the price although it seems not nearly as important an issue as it is in the UK.

2.28 A direct comparison of UK and US installation charges shows that installation is generally cheaper in the US than in the UK. This is especially evident for longer term contracts with some operators suggesting that they would be prepared to waive their connection charges for contracts of 2 years and over.

2.29 An initial comparison of published prices indicates that some US operators’ rental prices are higher than those applying to BT. Even on a very dense route (such as New York to Washington) Tarifica concluded that the list price of a 45 Mbit/s circuit ($66,372 per month) was considerably more than the list price of an equivalent distance 34 Mbit/s circuit with BT ($37,783 per month).

2.30 But these comparisons are almost meaningless as they take no account of discounts which can be extremely large in the US. Tarifica found, for example, that discounts available from a US operator on circuits for speeds up to 1.5 Mbit/s ranged from 11% to 52% while for speeds of 45Mbit/s and above discounts could exceed 80% for some types of customer. This contrasts with the UK where volume and term discounts from BT can result in a combined maximum discount of no more than about 35%.

2.31 The reason that the US has such large discounts is because US operators are able to negotiate deals on an individual basis with each customer. Within the UK, OLOs offer unpublished discounts and negotiate on an individual basis with each client. BT, however, is only able to offer published discounts under the terms of its Licence in order to ensure other operators can compete with BT.

2.32 It has been known for some time that these large discounts are a significant advantage to large companies in the US who have sufficient buying power to negotiate highly competitive deals for themselves. Such companies therefore find US prices to be cheaper than in the UK or the rest of Europe. What has been slightly less clear is how easy it is for smaller companies to negotiate similarly attractive deals.

2.33 The Tarifica analysis, however, appears to indicate that while the discounts are not of the same scale as for large companies they are still very significant in the US and, for most types of user, will pull the price down below levels they would pay in the UK.

2.34 It should however be noted that this conclusion is a generalisation and will vary according to the routes being considered as well as the user. It seems quite plausible, for example, that the price of a leased line on a very sparse non-competitive route in the US could be well above BT’s geographically averaged prices. In addition there may be plausible explanations for the price differential between the US and UK, such as economies of scale caused by the much larger volumes in the US, or different levels of service. This should not, however, detract from the fact that US private circuit prices do appear to be below prices in the UK.

2.35 In terms of European comparisons it is clear that BT’s prices for leased lines are broadly in line (or slightly cheaper in some instances) than prices charged by incumbent operators in other Member States. The UK’s relative position has worsened over the last couple of years as European operators have liberalised their markets which has resulted in some very significant price falls in a number of countries. However, these price falls were not unexpected given that prices in many of these countries were starting from a very high base.

2.36 As regards US comparisons, it is evident that US operator prices are lower than BT’s prices, particularly on highly competitive routes. The exact price paid in the US will however vary according to type of user and density of route as well as according to volume and length of contract.

Chapter 3

3.1 Oftel believes that, ideally, competition should provide the pressure for prices of leased lines to be reduced and for improvements in efficiency to be made and the cost savings to be passed on to users. Oftel would only wish to control prices directly where there was little or no prospect of competition.

3.2 This chapter sets out the appropriate market definitions and analyses the state of competition in each relevant market. It discusses the potential barriers to the development of competition and suggests the way forward.

3.3 This section will explain the relevant markets that Oftel believes are relevant in considering the state of competition in national leased lines. The approach adopted here is consistent with the approach set out in the Office of Fair Trading guidelines on market definition (OFT (1999) Market definition, OFT403).

3.4 Private circuits or leased lines are generally made up of a trunk segment (for example, digital trunk circuits linking two cities) and terminating segments (which link the customer premises to the operators’ trunk network). While a leased line may pass through a local or trunk exchange building they do not pass through the actual Digital Local Exchange (DLE) or Digital Main Switching Unit (DMSU) switches. It is important to note that terminating segments are distinct from local ends (see diagram below). A local end is a connection from the customer's premises to the local exchange level of an operator's network. Terminating segments are, in general, connections from the customer's premises to the DMSU level of an operator's network.

A --------Lxa -----------DMSU ------------------------------------DMSU -----------LXb ---------B

{ local end }-------------{ main link }-----------------{ local end}

{ terminating segment }--------------------{ trunk segment }----------------------{ terminating segment }

3.5 In the past, Oftel has generally examined the markets for the provision for end to end private circuits. In the previous Statement, Oftel noted that ‘other operators may not always have sufficient infrastructure in place to offer customers a full end-to-end leased line service and will therefore need to purchase the 'final mile' circuits from other operators’. Having recognised that there are differences in ability to provide the different parts of a private circuit, Oftel considers it appropriate to investigate the possibility of defining separate economic markets on this basis. Oftel believes that these market definitions may be able to capture more fully the dynamics of competition for the provision of private circuits.

3.6 In general, operators providing a leased line service to end-users will need to either purchase or build terminating segments. If operators build terminating segments they are likely to incur high sunk costs. Sunk costs include costs of digging and ducting fibre/copper wire from a customer’s premise to the operators’ trunk network. Costs are ‘sunk’ because they are unrecoverable if the operator were to exit the market.

3.7 Oftel takes the view that there is a case for identifying separate markets for the trunk segment and the terminating segments of a private circuit. It is necessary to consider the extent to which prices of the terminating segments are constrained by the availability of substitutes. A trunk segment is not a substitute for a terminating segment – in fact, they are complements ie they both need to be purchased together in order to provide an end to end leased line. It is also necessary to consider the extent to which operators who are providing a similar product (eg trunk segments) could switch into supplying terminating segments. As stated previously, sunk costs for the provision of terminating segments are quite large. This means that on the whole, entry barriers faced by an operator are high and there is unlikely to be much substitution from providers of trunk segments switching into the provision of terminating segments. Thus, there would appear to be distinct markets for provision of trunk segment and the provision of terminating segments.

3.8 Terminating segments may be used to provide services other than end to end leased lines. For example, operators may use terminating segments to provide Frame Relay, ATM and Internet access services.

3.9 There may be some scope for arguing that each terminating segment should be treated as a separate market. It can be claimed that the connection of a terminating segment into one customer's premises is not a substitute for a connection into another customer's premises. This implies that the prices charged in one area may not be constrained by the prices charged in another area. However, BT’s geographically averaged pricing structure (outside CLZ) affects the competitive pressures faced by all firms in the market. All operators are constrained by BT’s prices. This means that a price rise above competitive levels of an individual terminating segment would be rendered unprofitable because of BT’s averaged pricing across all its terminating segments. Thus, it would appear that all terminating segments together constitute a distinct economic market.

3.10 In 1997, Oftel made a distinction between 2Mbit/s and above (high) and 64kbit/s and below (low) capacity circuits. There was a distinction in the way such circuits were delivered with low speed circuits using copper pairs in the local end and high speed using fibre. However, given the wider use of HDSL over copper to deliver 2Mbit/s circuits in less dense areas, Oftel would suggest that 2Mbit/s circuits should now be classified as low capacity circuits.

3.11 Oftel considered whether there should be distinct markets for circuits of different capacity. Oftel’s approach was to obtain information on whether a supplier of high / low capacity terminating segments could easily switch into supplying low / high capacity terminating segments if the price of the latter product(s) rose above competitive levels.

3.12 Given the high sunk costs described earlier, a supplier of high capacity terminating segments would be unlikely to switch into supplying low capacity terminating segments. It would be unlikely to be able to justify the large sunk costs incurred in order to supply low capacity terminating segments. This would indicate that there are likely to be distinct economic markets for high and low capacity terminating segments.

The market for trunk segments of private circuits

3.13 There may be some scope for arguing that each trunk segment route should be treated as a separate product market. Substitution by consumers is unlikely on trunk segments, since, for example, a link connecting the Manchester-London route is not a substitute for a Liverpool-London route. An operator on the former route who wished to offer a service on the latter route would need to build a new trunk segment either connecting Liverpool and Manchester or Liverpool and London. This is more likely to occur on routes that are going to generate large volumes of traffic. On a route that connected London to a rural area, where there may not be large volumes of traffic, an alternative supplier may not be attracted into that market because of the large sunk costs it would need to incur to set up the trunk segment.

3.14 However, the existing pricing structure by BT, the largest player in the market, affects the competitive pressures faced by firms in the market. Specifically, BT's geographically averaged prices (outside the CLZ area) means that its ability to raise prices on low volume routes (where it might face less competition) is to a certain extent constrained by the entry it might incur on high volume routes. This pricing strategy implies that currently, the relevant product markets are not based on individual routes but are likely to include all trunk segments.

3.15 It should also be considered whether there should be distinct markets for trunk segments of different capacity. This can be tested by considering whether a supplier of high / low capacity trunk segment could easily switch into supplying low / high capacity trunk segments if the price of the latter product(s) rose above competitive levels.

3.16 The ease with which a supplier of high capacity, trunk segment circuits could switch into supplying low capacity trunk segment circuits would depend on three main factors:

• the existence of suppliers of different capacity on the same route;

• the volume of traffic on the route; and

• pricing structures.

3.17 If there were suppliers already offering high capacity circuits on the routes in which the price of low capacity circuits have been increased, it would be relatively easy for the high capacity supplier to switch into supplying low capacity circuits. Thus, the potential for substitution by high capacity suppliers would imply that the relevant market for trunk segments of private circuits included both high and low capacity circuits.

3.18 If there were no suppliers offering high capacity circuits on the trunk segment routes where the prices of low capacity circuits have been increased, then the extent to which a supplier of a similar service (eg supplier on a different route) could switch into offering services on that route would depend on the volume of low capacity traffic on that route. An alternative supplier would set up a trunk segment circuit on the route if he could generate a large enough volume of traffic to justify the large sunk costs. Thus, supply side substitution or entry would only occur on routes where there was sufficient volume. On these types of routes the relevant economic market definition would include both high and low capacity circuits.

3.19 The implication is that if there were no high capacity suppliers and there was not a large volume of traffic on the route then supply side substitution would be less likely to occur. This is because of the large sunk costs involved in supplying low traffic volume routes. This suggests that there may be a case for identifying separate markets for high and low capacity circuits. However, BT has a nationally averaged pricing structure (excluding the CLZ area). It is thus unable to focus price rises on the low volume routes where it is less likely to face competitive entry. BT might still choose to raise prices of low capacity trunk segments, suffer loss of market share on routes where there were high traffic volumes and gain higher revenues from low capacity, low traffic volume routes. However, its ability to do this is restrained by the extent of loss in revenue it might suffer on low capacity, high traffic volume routes. The implication of this is that prices of BT's low capacity trunk segments are to some extent constrained by threat of entry from high capacity suppliers. This pricing structure suggests that to the extent that the profitability of BT’s price rises is constrained by threat of entry on high volume routes, low and high capacity trunk segments are likely to be in the same market.

3.20 BT’s pricing structure affects the nature of competition in the market. BT charges a nationally averaged price for leased lines in all geographic areas within the UK, except where both ends of circuits are within CLZ. Therefore, other leased lines suppliers face national (excluding CLZ) competitive pressures. CLZ prices, however, are lower, in general, than prices in other parts of the UK. The prices in CLZ area do not appear to constrain the prices that suppliers of leased lines can charge outside the CLZ area. This implies that there are two relevant geographic markets: CLZ and rest of the UK.

3.21 This section explains Oftel’s view of the state of competition in each of the relevant product markets based largely on information gained during the review. To summarise, the relevant product markets are:

• Low capacity (2Mbit/s and below) terminating segments;

• High capacity (above 2Mbit/s) terminating segments; and

• Trunk segments.

3.22 In each market Oftel considers to what extent BT has and is likely to continue to have market power. The assessments include available evidence on market shares, presence / absence of competitors, level of entry barriers and relevant regulatory framework.

3.23 Available evidence suggests that BT has a very high market share in the supply of low capacity terminating segments in CLZ and in the rest of the UK. This is based on information for end to end private circuits and the assumption that BT would self supply its terminating segments. In terms of 1998/9 revenues, BT has a share of approximately 96% of the market segment for below 2Mbit/s private circuits. Information is not available on BT’s share of 2Mbit/s for 1998/9, although available data suggests that BT’s market share is greater than 70%.

3.24 Market shares of other operators for low capacity terminating segments appear to be fairly low. Very few operators offer circuits at below 2 Mbit/s. A slightly greater number of operators are more likely to offer terminating segments at 2Mbit/s. Primarily these are cable operators in their franchise areas, some operators in CLZ area and regional operators in their regions. But their share of this market segment is relatively low.

3.25 Entry barriers are high for low capacity terminating segments. The large sunk costs necessary to set up terminating segments implies that new entry is unlikely at low bandwidths since operators are unable to recover these costs over a reasonable time period.

3.26 On the basis of high market shares and high entry barriers, it would appear that BT has considerable market power in the supply of low capacity terminating segments. Currently, BT operates under a retail price cap (RPI+0%) for 64kbit/s leased lines. The regulatory framework does not oblige BT to offer an interconnect product and/or offer wholesale prices. This is primarily because of Oftel’s policy decision to promote infrastructure competition in provision of leased lines.

3.27 For the most part OLOs purchase low capacity terminating segments from BT. A terminating segment is used as an input not just into the provision of leased lines but also to provide other products, for example, internet access. Since BT is not obliged to offer an interconnect product, an OLO has to purchase a retail leased line from BT in order to connect a customer to their network at 64kbit/s.

3.28 At 2Mbit/s, BT has made a wholesale product termed a partial private circuit (‘PPC’) available to OLOs. In theory, these could be purchased by OLOs to provide the terminating segments. However, these are also priced on a retail basis ie price of PPC equals the retail price of a private circuit minus the cost of one local end and minus 50% of the fixed charge for the main link. But take up of this product has been very low. OLOs have made representations to Oftel claiming that the discount policy operated by BT has prevented the take up. Their reasons for non-take up mainly include:

• higher level of discounts available for retail leased lines relative to those for partial private circuits;

• no clear migration path from purchasing retail leased lines to purchasing partial private circuits; which means that if OLOs migrated to PPCs they would suffer penalties for not meeting their volume commitments for leased lines; and

• no discount pool which combines retail leased lines and partial private circuits.

3.29 Although Oftel’s policy of encouraging infrastructure competition has met with success at some levels, such as competition in the provision of trunk segments, its success has been limited in relation to competition for low capacity terminating segments. There has been limited entry into this market. Given BT’s continued market power, there is a case for reviewing whether an interconnection product should be offered at an interconnect price. Oftel therefore, believes that it is appropriate to review the requirement to supply low capacity terminating segments, including considering the appropriateness of the existing retail price control on 64kbit/s private circuits.

3.30 This is an area of the market that is expected to grow very rapidly over the next few years. To date, the volume of private circuits at these bandwidths has been very limited. Market share information is very thin for high capacity terminating segments. If market share is measured in terms of volume, it is likely that BT may not have considerable market power in the market. However, BT’s market share in terms of revenue is likely to be much higher. Lack of data means that it is very difficult to estimate reasonably robust market share data. It is also difficult to obtain a breakdown on a geographical basis. It would appear that competition is greater within the CLZ area relative to the rest of the UK.

3.31 In general, entry barriers are less acute than for lower capacity terminating segments. But the amount of local infrastructure investment for provision of leased lines is still limited (although a few OLOs are considering limited local infrastructure investment for leased lines).

3.32 In their representations to Oftel, OLOs indicated that they still often need to buy terminating segments from BT. OLOs expressed concern that it is sometimes not economically viable for operators to pay the large sunk costs for digging and ducting. Sunk costs can be recovered either through high up front charges so that they are recovered early in the contract or by commitment to long term contracts so that they can be recovered over a longer period of time. OLOs have suggested that neither of these options is favoured by customers and that as a result they face difficulties in incurring the large sunk costs of digging and ducting to build terminating segments. OLOs also claimed that it was difficult for operators to compete with BT because of lead-time considerations. Often, in order to provide an immediate service, competitors needed to buy a terminating segment from BT who already had ducting in place. It was claimed, therefore, that in some cases it would not be feasible for OLOs to build high capacity terminating segments even where it might be justified on the basis of cost.

3.33 Oftel believes that it is possible that OLOs may be relying on BT for provision of terminating segments at 8Mbit/s. This derives from the fact that BT enjoys economies of scope in duct network between leased lines and other uses so it will always find 2-8Mbit/s circuits more economic to deliver over fibre (or even copper) than an OLO. However, Oftel has received some indication that OLOs are less likely to need to rely on BT for private circuits of 34Mbit/s and above. At these higher bandwidths it would appear that OLOs find it economic to offer bespoke pricing for circuits such that they adequately recover their sunk costs. Moreover, some evidence suggests that only a tiny minority of its 34Mbit/s and above private circuits are sold to OLOs. Thus, there is some indication that OLOs have been able to compete with BT in the provision of terminating segments of 34Mbit/s and above.

3.34 If OLOs were to purchase high capacity terminating segments from BT, they buy it at a retail price. BT does not offer a partial private circuit for high capacity terminating segments. BT has argued that the limited take up of the 2 Mbit/s partial private circuit has indicated that there is no demand for PPCs. However, Oftel’s discussions with OLOs have suggested that take up has been limited due to the prices and discount schemes under which the 2Mbit/s PPC is offered. The regulatory framework does not require BT to provide an interconnect product at a wholesale price. The purpose of this was to encourage network build and network competition.

3.35 Oftel believes that competition may be developing at different rates within the market for high capacity terminating segments. There are prospects for competition especially for circuits of 34Mbit/s and above. Overall, it appears that BT may have some market power in the provision of high capacity terminating segments. However, the information available relating to this market is limited. As part of the proposed review, Oftel will consider further the extent of BT’s market power, prospects for further infrastructure investment and the policy implications for high capacity terminating segments.

3.36 Available evidence suggests that BT has a high market share in the provision of trunk segments. This is derived from data on end to end leased lines, where 1998/9 revenue figures suggest that BT has a market share of around 83%. Unfortunately, it has been difficult to break this down on a geographical basis. However, it would appear that BT faces greater competition in the CLZ area relative to the rest of the UK.

3.37 A number of OLOs compete with BT in the provision of trunk segments. This is shown in the table below.

Note : A number of the regional operators are expanding the geographical scope of their provision.

3.38 It would appear that entry barriers are relatively low on major metropolitan routes. However, on non-major metropolitan routes, entry barriers are likely to be higher due to lower volumes of traffic.

3.39 There has been significant network investment by operators and Oftel has been informed that some companies plan new infrastructure investment. To this extent Oftel’s policy of encouraging network competition has clearly been successful.

3.40 Overall, it appears that competition varies between routes and competition is more likely to be effective on metropolitan routes. However, BT’s geographically averaged prices prevent it from raising prices on routes where it faces less competition and thereby limits its ability to charge excessive prices.

3.41 BT is subject to a price cap on 64kbit/s private circuits. However, given that competition and geographically averaged prices may be constraining BT’s ability to raise prices on low capacity trunk segments, there is a case for removing the current retail price cap for trunk segments.

3.42 The evidence on the state of competition is mixed. Competition is increasing in the provision of trunk segments. However, Oftel has concerns about the provision of terminating segments, especially at lower bandwidths. Oftel is concerned that the regulatory framework in relation to the provision of terminating segments should not act as a barrier to the development of competition. Oftel proposes to conduct a major review of the regulatory framework to ensure that it continues to be appropriate. Further details of this review are given in the next Chapter.

4.1 Oftel’s current regulatory framework specifically relating to national leased lines has two principal aspects: a retail price control of RPI+0% for private circuits at 64kbit/s and below; and no obligation on BT to offer a wholesale leased line service to other operators at interconnect prices. There are also general regulatory rules relevant to the provision of leased lines, in particular price publication and notification. This framework was put in place following Oftel’s review of regulatory controls on BT’s prices in 1996, (Pricing of telecommunications services from 1997, June 1996) and so reflected Oftel’s assessment of the prospects for infrastructure competition, for competition in the provision of leased line services and the state of technology at that time.

4.2 Oftel’s recent investigation into leased lines, and other developments, have revealed that the state of competition in markets and technology has developed. Oftel now has concerns as to whether the current framework continues to be appropriate, in particular whether the mix of regulation and de-regulation is right.

4.3 As discussed in Chapter 2, Oftel’s review of international price comparisons suggests that, while the UK is broadly comparable with European countries, its position has been eroded and in relation to the US it appears that for most types of user US price levels are below those in the UK.

4.4 As explained in Chapter 3, Oftel has considered further its concerns identified in the previous Statement relating to ‘final mile’ circuits. It has carried out a more detailed market analysis of the different markets in order to obtain a deeper understanding of dynamics of competition relating to the provision of end to end leased line services. It has identified separate markets for trunk segments and terminating segments. The review found that there are competition concerns relating to the terminating segments. In this market BT has, and seems likely to retain, market power (as explained in Chapter 3 the extent of the market power varies with capacity), which suggest there may be a need for targeted regulation. By contrast in the market for trunk segments the review found competition was developing and there do not appear to be competition concerns which suggests that there may be scope for de-regulation.

4.5 The further review of competition also found that since the current regulatory framework was put in place considerable investment in infrastructure has occurred in the provision of trunk capacity. However, the building of local infrastructure for the provision of leased lines has been more limited.

4.6 There are also other developments which have given rise to concerns with the current regulatory framework. Oftel is aware that BT’s market power in termination segments has implications for the level of competitiveness in the provision of services other than end to end leased lines such as frame relay services, internet access services and ATM services.

4.7 Also, responses to Price Control Review: Future Developments in the Competitiveness of UK Telecoms Markets, published in July 1999, which was the first stage of consultation in Oftel’s review of retail and network price controls, revealed concerns by users and operators about the pricing of leased lines. In particular, issues about the competitiveness and pricing of the terminating segments of private circuits and the relative price levels of short and long distance circuits were raised.

4.8 Oftel has therefore decided that it is now appropriate to launch a full review of its regulatory framework relating to national leased lines. Rather than carry out this review as part of the Price Control Review, Oftel has decided to set up a specific project to focus on national leased lines. The plan is to complete the review and publish a Statement by October 2000.

4.9 In a parallel development, the European Commission has analysed the competition concerns relating to the provision of leased lines in a way which is broadly similar to that presented by Oftel in this Statement. In the Directorate General Information Society’s analysis, which underlies the Commission’s recent recommendation on leased line interconnect pricing (see the Information Society DG Working Document, ‘Commission Recommendation on leased line interconnect pricing’ 31 August 1999), it has focused on the market power of incumbents in the provision of short distance private circuits which it believes new entrants need to purchase in order to offer end to end leased lines. Oftel’s Review of its Regulatory Framework will address the implications of the Commission’s Recommendation for the UK.

4.10 This new Review is a major project for Oftel. It will take forward the competition concerns and scope for de-regulation outlined in this Statement, all issues related to the pricing of national leased lines raised in the context of the Price Control Review and those raised by the Commission’s Recommendation on leased line interconnect pricing. The project will carry out a detailed investigation into BT’s costs and prices for the provision of termination segments. This will, however, require analysis of prices and costs of trunk segments as well. In this regard, it is anticipated that Oftel will work with consultants to develop a suitable model to analyse prices and costs for private circuits. The project will also consider to what extent it is appropriate to deregulate the trunk segment market, for example in terms of price regulation and price publication. It is anticipated that discussions with operators will be held, in particular to assist Oftel in understanding their plans for infrastructure development and their requirements in relation to a wholesale interconnect leased line product for the provision of end to end leased lines and in relation to the provision of other services such as Frame Relay, ATM and internet access.

4.11 The main issues and questions which the Review will consider are as follows:

1. If BT were obliged to offer an interconnect leased line product at interconnect prices, what would be the implications for investment in network infrastructure? Would the implications vary with: (a) the interconnect product offered and (b) the bandwidth concerned?

2. If BT were obliged to offer an interconnect leased line product at interconnect prices, what would be the implications for retail leased line prices? Would rebalancing occur? For example, would prices for longer distance circuit prices fall but shorter distance circuits rise? Would prices for high bandwidths fall but lower bandwidths rise?

3. If BT were obliged to offer an interconnect leased line product at interconnect prices, what would be the implications for the development of competition in other markets, other than end to end leased lines, for which leased lines constitute an input such as ATM, Frame Relay and internet access? Would the implications depend on the interconnect product offered?

4. If BT were to offer an interconnect leased line product, what would be the appropriate product? Is the existing 2Mbit/s Link to a Point of Connection product (a Partial Private Circuit) an appropriate model?

5. If BT were to offer an appropriate interconnect leased line product, would there be demand from other operators at all analogue and digital bandwidths?

6. Should the regulatory rules vary with bandwidths? If the local loop is unbundled in line with Option 2 (‘Access to bandwidth: Proposals for action’, July 1999) how would this affect the need for an interconnect leased line product at lower bandwidths?

7. Should the same regulatory rules apply throughout the UK? Should the CLZ have a different set of regulatory rules to the rest of the UK?

8.   Should the existing price cap for analogue and 64kbit/s for trunk segments be removed given the increasing competition which appears to be developing in this market ?

9.    Should BT be allowed to offer bespoke deals / unpublished discounts?

C. Market definitions

10. Are the market definitions, set out in Chapter 3, an appropriate way to analyse competition in the provision of leased line services?

11. Is the distinction between low and high capacity set at the correct level? Should high capacity only relate circuits above 34 Mbit/s?

4.12 Oftel will continue to monitor closely international price comparisons for leased lines both in relation to other European countries and the US. As well as subscribing to various Consultancy reports, Oftel intends to investigate further the pricing structures that exist in a few specific countries that are felt to be of most interest. This should enable Oftel to interpret more accurately price comparisons between these countries and obtain a better understanding of international prices.

4.13 DG Competition has launched a sector inquiry under Regulation 17/62 into the telecommunications sector. The first phase of this inquiry will relate to leased lines. Oftel will be participating in and contributing to this inquiry.

Comments on this Statement are invited, particularly in relation to the issues and questions outlined in paragraph 4.11 and the account of the state of competition given in Chapter 3. Oftel is seeking the views of operators, service providers and consumers by Monday 31 January 2000. Comments should be made in writing and sent to:

Tim Cross
Oftel
50 Ludgate Hill
London
EC4M 7JJ

Tel: 0171 634 8798
Fax: 0171 634 8772

Alternatively comments may be subbmitted by email.

Written comments will be made publicly available in Oftel’s Research and Intelligence Unit except where respondents indicate that their response, or parts of it, are confidential. Respondents are therefore asked to separate out any confidential material into a confidential annex which is clearly identified as containing confidential material. In the interests of transparency, respondents are asked to avoid confidentiality markings wherever possible. Appointments to view written comments in Oftel’s Research and Intelligence Unit, which must be made in advance, can be arranged by ringing: 0171 634 8761.

Oftel would like to set up a link between this Consultative Document and any responses placed on respondents own Internet pages. Please contact Lauren Ryner at Oftel on 0171 634 8753 or by e-mail  to arrange this. Confidential responses should not be sent via the Internet.

Copies of the full Consultative Document are available on disk.

The Summary is available in large print, Braille, and tape formats.

Please contact the Oftel Research and Intelligence Unit on 0171 634 8761, or by e-mail , or call textphone 0171 634 8769 for more information.

ATM service – data services using Asynchronous Transfer Mode technology, such as BT’s Cellstream service or Energis’s Cellconnect

Central London Zone (CLZ) pricing – BT offers reduced charges for some leased lines if both ends are within the 0171 exchange area.

DSL technology – digital subscriber line technology.

DSMU (Digital Main Switching Unit) – a trunk exchange primarily used for connecting calls between DLEs.

DLE (Digital Local Exchange) – the telephone exchange to which customer are directly connected.

Frame Relay service – a packet switched data service providing for the interconnection of Local Area Networks and access to host computers at up to 2 Mbit/s.

Leased line – A permanently connected communications link between two premises dedicated to the customers’ exclusive use. Also known as a private circuit.

Local end circuits – The leased line between a customer’s premises and the local exchange

Main link – The leased line between the local exchanges of the customer’s premises.

Mbit/s – Mega (million) bits per second. A measure of speed of transfer of digital information.

‘Multi-vendor’ circuits – A leased line provided to a customer which uses the network infrastructure of more than one operator. eg –one local end and a main link from operator A and a final mile circuit rented by operator A from operator B.

Partial private circuit – BT’s Link to a Point of Interconnection product. This is a private circuit from a customer’s premises to point of interconnection between BT and another operator’s network. It is only available at 2Mbit/s to be purchased by PTOs.

Own-exchange circuit – A leased line linking two premises connected to the same local exchange.

Terminating segment – The communications link required by an operator to complete the connection from the operator’s network (usually at DSMU level) to the customer’s premises.

Trunk segment – The communications link connecting trunk exchanges.