Direction under the provisions of Regulation 6(3) of the Telecommunications (Interconnection) Regaulations 1997 of BT's Retail Uplift charge for calls to operators' Number Traslation Services from 1 April 2001

Contents

The Final Direction

Explanatory Memorandum

Chapter 1 Summary

The decision

Chapter 2 Background

Chapter 3 Oftel’s methodology for calculating the Retail Uplift
Historic methodology
Current methodology

Details of the methodology
Baseline
Methodology to update the cost allocation to the current year
Volume adjustment
Efficiency adjustment
Composition of BT’s retail costs
Updating the charge

Chapter 4 Comments received to the draft proposals and Oftel’s responses

BT

Process
Methodological validity
Efficiency gains
Use of 1996 ‘CVR’
Recovery of costs
Incorrect 2002/03 charge
Marketing & Sales costs

Other Operators

Freephone calls
Variation in the retail uplift from year to year
Lagged data
Retrospection
ROCE and efficiency
Bad debt
Marketing & Sales costs
Implementation of directions

 Chapter 5 The Director’s final decision

The relevant market and the position of BT
Exceptional circumstances
The Director’s final decision
Implementation of directions

Annex 1 Why Marketing and Sales costs should be included in the Retail Uplift
Calculation of efficiency savings in retail costs

Direction under the provisions of Regulation 6(3) of the Telecommunications (Interconnection) Regulations 1997 of BT’s Retail Uplift charge for calls to operators’ Number Translation Services from 1 April 2001

WHEREAS:

A) The Secretary of the State granted to British Telecommunications on 22 June 1984 a licence ("the Licence") under section 7 of the Telecommunications Act 1984 ("the Act") for the running of the telecommunication systems specified in Annex A to the Licence;

B) By virtue of Section 109 of, and paragraph 20 of Schedule 5 to, the Act, the Licence has effect as if granted to British Telecommunications plc ("BT");

C) BT, in accordance with Condition 45 of its licence, has entered into interconnection agreements with a number of Operators. Those interconnection agreements include provisions concerning Number Translation Services ("NTS"). The term "Operators" in this direction shall refer to those operators that have entered into an interconnection agreement with BT; 

D) By way of a determination entitled Interim Charges for BT’s Initial Standard Services for the year ending 31 March 1996 the Director General of Telecommunications ("the Director") determined a formula for NTS services where the call originated on one operator’s network and terminated on another. This formula ("the NTS Formula") may be summarised as follows:

Originating Network Operator ("ONO") keeps P – D + C

Terminating Network Operator ("TNO") keeps D – C

Where

E) "C" as described in the NTS Formula continues to apply (for the purpose of assessing BT’s retention) in accordance with, inter alia, the November 1999 Direction concerning BT’s NTS Conveyance and the December 1999 Statement on the Relationship between Interconnection Charges and Retail Prices for Number Translation Services.

F) The Director has previously reviewed and set the NTS Retail Uplift for the financial years commencing 1 April 1999 and 1 April 2000.

G) The Telecommunications (Interconnection) Regulations 1997 ("the Regulations") inter alia implement Directive 97/33/EC on interconnection in telecommunications with regard to ensuring universal service and interoperability through application of the principles of open network provision ("the Directive");

H) Regulation 6(1) of the Regulations provides that the Director shall encourage and secure adequate interconnection in the interests of all users and that he exercises his functions in a way that provides maximum economic efficiency and gives maximum benefit to end-users having regard to the matters set out in Regulation 6(1)(a) to (g) of the Regulations.

I) Pursuant to Regulation 6(3) of the Regulations the Director may intervene at any time, in order to make a direction specifying issues which must be covered in an interconnection agreement, or to make a direction that specific conditions be observed by one or more parties to such an agreement. The Director may in exceptional circumstances make a direction that changes be made to interconnection agreements already concluded where it is justified to ensure effective competition or interoperability of services for users or both; 

J) A draft of this direction and the explanatory memorandum was published on 2 January 2003 and comments invited by 30 January 2003;

K) Comments were received from BT, Easynet, Cable and Wireless and Energis and the main points made by those who responded are summarised in Chapter 4 of the explanatory memorandum which accompanies and is published with this direction.

THEREFORE

Pursuant to the provisions of Regulation 6(3) of the Telecommunications (Interconnection) Regulations 1997, the Director General makes the following direction: 

1. The NTS Retail Uplift applied by BT to assess its NTS conveyance charge, namely "C" as described in recital D, from 1 April 2001 until 24 July 2003 shall be as follows:

(i) from 1 April 2001 to 31 March 2002 inclusive:

An average retail uplift charge of 0.1911ppm, which gives rise to:

(ii) from 1 April 2002 to 24 July 2003 inclusive:

An average retail uplift charge of 0.2044ppm, which gives rise to:

2. Any amount payable by BT to an Operator as a result of this direction should be paid together with interest calculated in accordance with Clause 13.13 of their interconnection agreement.

3. Any amount payable by an Operator to BT as a result of this direction should be paid together with interest calculated in accordance with Clause 13.13 of their interconnection agreement.

4. In accordance with paragraphs 5.8 to 5.14 of the explanatory memorandum which accompanies and is published with this direction, BT shall alter the Carrier Price List so that it accords with this direction.

5. The terms defined or described in the recitals to this direction shall have the meaning so defined or described. All other words or expressions used in this direction shall have the same meaning as in the Directive, the Regulations, the Act or the Licence as appropriate. 

Heather Julie Clayton

Director of Investigations

A person authorised under Paragraph 8 of Schedule 1 to the Telecommunications Act 1984

25 March 2003

Explanatory Memorandum

Chapter 1

Summary

1.1 The Director General of Telecommunications ("the Director") has issued a final direction, in accordance with the provisions of Regulation 6(3) of the Telecommunications (Interconnection) Regulations 1997 ("the Regulations"), to review the methodology for setting BT’s Retail Uplift within its regulated Number Translation Services (NTS) call origination charge and to set BT’s NTS Retail Uplift for the following periods:

• from 1 April 2001 to 31 March 2002; and
• from 1 April 2002 to midnight on 24 July 2003:

1.2 Oftel has devised a methodology for calculating an uplift allowance for retail costs consistent with the principles of cost orientation. In order to meet this objective, the methodology re-visits the initial allocation of costs that had, in the past, been implicitly allocated to NTS services. This allocation is then updated over time to the current year, in order to reflect economies of scale due to volume growth, cost reduction due to BT’s efficiency in retail costs and inflation adjustments.

1.3 The Director published a draft direction and explanatory memorandum on 2 January 2003. In reaching his final decision the Director has considered the comments submitted in response to the draft direction and the requirements set out in Regulation 6(1) of the Regulations, in particular the need to stimulate a competitive market for the provision of services accessed by NTS calls.

The decision

1.4 Accordingly, the Director has decided that, with effect from 1 April 2001 until 31 March 2002 inclusive, the retail uplift for relevant retail costs applied by BT to give its NTS Conveyance charge should be:

An average retail uplift charge of 0.2044ppm, which gives rise to:

1.5 The Director has also today issued three related documents:

• a final direction re-amending a direction made on 5 April 2002 which set BT’s NTS retail uplift charge for the period from 1 April 2000 until 31 March 2001 in order to correct the same errors entitled "Re-amended Direction under the provisions of Regulation 6(3) of the Telecommunications (Interconnection) Regulations 1997 of BT’s retail uplift charge for calls to operators’ number translation services from 1 April 2000";

• a final direction to amend a direction made on 19 September 2002 which set BT’s NTS Retail Uplift charge for the period from 1 April 1999 until 31 March 2000 entitled "Amended Direction under the provisions of Regulation 6(6) of the Telecommunications (Interconnection) Regulations 1997 resolving a dispute between Energis and BT concerning BT’s method of calculating its NTS retail uplift charge since April 1997"; and

• a final direction resolving the dispute between BT and CWC relating to the surcharge for PRS bad debt entitled "Direction under the provisions of Regulation 6(6) of the Telecommunications (Interconnection) Regulations 1997 resolving a dispute between Cable & Wireless Communications (Mercury) Limited ("C&W") and British Telecommunications plc ("BT") over a bad debt surcharge relating to calls to Premium Rate Services".

Chapter 2

Background

2.1 In previous determinations of NTS charges Oftel has given a detailed description of the NTS revenue sharing arrangements and how they were arrived at. Most operators are now familiar with this narrative and it is not, therefore, repeated in this document. Anyone wishing to read the detailed description can refer to Oftel’s earlier NTS directions (also known as determinations) which can be found on Oftel’s website.

2.2 Oftel first established the principle of a supplemental charge to meet BT’s relevant retail costs in providing access to NTS services, in the first ICAS Determination of Interim Charges for BT’s Initial Standard Services for the year ending 31 March 1996 ("the 1996 Determination") published on 30 January 1996. The original charging methodology for calculating BT’s retail uplift has become known to the industry as the ‘broad brush’ methodology.

2.3 In 2001, Oftel reassessed the methodology for calculating BT’s NTS retail uplift and a draft direction proposing to set the retail uplift for NTS calls for the year 1 April 2000 to 31 March 2001 was issued in October 2001. A final direction was issued on 28 March 2002 and amended on 5 April 2002. This direction is referred to as the "April 2000 direction". A further amendment has been made to the April 2000 direction today.

2.4 On 1 May 2002 Energis gave notice to the Director of its intention to appeal the April 2000 decision. On 30 October 2002 Energis withdrew its appeal in the light of its understanding of Oftel’s work programme for NTS.

2.5 On 19 September 2002 Oftel issued the "Direction under the provisions of Regulation 6(6) of the Telecommunications (Interconnection) Regulations 1997 resolving a dispute between Energis and BT concerning BT’s method of calculating its NTS Retail Uplift charge since April 1997". This followed a referral, on 8 February 2002, by Energis of a dispute with BT over the recalculation of the retail uplift charge back to 1997. The decision in this direction was to set BT’s NTS Retail Uplift charge from 1 April 1999 and is therefore referred to as the "April 1999 direction".

2.6 In the April 2000 direction Oftel stated its intention at that time to consult on a mechanism for updating BT’s retail uplift using a price cap. Oftel intended that this price cap should apply from 1 April 2001. Oftel put some of this work on hold pending the outcome of the Energis appeal. Oftel has recently announced a series of market reviews in preparation for the implementation of the new EU Directives on midnight 24 July 2003. Regulation in this new regime must, in most cases, follow a market review. Given the delay to Oftel’s work on the retail uplift and the requirements of the new EU Directives, Oftel considers that it is not appropriate to proceed with a consultation on a price cap under the current regime. Instead, Oftel proposes to proceed by way of this final direction to set the retail uplift for the remaining period from 1 April 2001 up to 24 July 2003. The means of setting the charge under the new regime will be considered within the market review process.

2.7 At the same time as publishing this final direction, the Director has also published the Final Direction under the provisions of Regulation 6(6) of the Telecommunications (Interconnection) Regulations 1997 resolving a dispute between Cable & Wireless Communications (Mercury) Limited ("C&W") and British Telecommunications plc ("BT") over Premium Rate Service calls bad debt surcharge.

2.8 In considering the dispute about the PRS bad debt surcharge, the Director discovered an error in the calculation of the NTS Retail Uplift in the April 1999 and April 2000 Directions. Some PRS bad debt costs were inadvertently included in the data underlying the calculations. The Director has therefore today published amendments to the April 1999 and April 2000 retail uplift Directions. These amendments are final amendments and are published together with this document and the final PRS bad debt decision. This document should therefore be read in conjunction with the explanatory memoranda that accompany those final amended directions and the final direction resolving the dispute about PRS bad debt.

Chapter 3

Oftel’s methodology for calculating the Retail Uplift

Historic methodology

3.1 Prior to April 2000, the retail uplift was calculated using the "broadbrush" methodology set by the 1996 Determination. This entailed deriving the charge by estimating an allowance for retail costs from BT’s annual accounting data on network and retail costs. Under the broadbrush approach, a ratio was derived by dividing BT’s total retail costs by BT’s total network costs. This percentage ratio was multiplied by the charge for BT’s single tandem network costs. Operators raised concerns and disputes about the simplicity of the broadbrush methodology. There are two main reasons why Oftel departed from this broadbrush methodology in its recent directions:

• the broadbrush methodology does not explicitly identify retail costs incurred in providing an NTS call origination service but uses a simple ratio of total retail and network costs. In reality, retail costs are not dependent on conveyance costs; and

• the broadbrush methodology was devised at a time when BT’s accounts were prepared on the basis of Historic cost Accounting ("HCA"). After August 1998, BT moved from HCA to Current Cost Accounting ("CCA"). This change leaves the broadbrush methodology open to a number of different interpretations, with no obvious way to resolve these uncertainties. A summary of these difficulties can be found in paragraphs 5.15 to 5.17 of the April 1999 direction.

Current methodology

3.2 Given the inadequacies of the broadbrush methodology, Oftel devised a new methodology based on relevant retail costs. Oftel’s methodology starts with an initial cost allocation (established during the 1996 retail price control review ("1996 PCR")). This allocation is then updated over time using up to date information for the relevant year for which the NTS retail uplift charge is being considered. The update takes into account estimated economies of scale due to volume growth, cost reduction due to BT’s efficiency in retail costs and inflation adjustments.

3.3 In BT’s Financial Statements, retail costs caused by NTS calls are mainly located in the Local and National call category of BT’s regulatory accounting business, the Retail Systems Business ("RSB"). The relevant retail costs in those categories are common between geographic and NTS call types. (see paragraphs 3.18 to 3.23 and Annex 1 for more details). Consequently, the relevant retail costs have been allocated between geographic and NTS calls according to relative volumes. Local and National calls, which are geographic calls, were regulated under the 1996 Price Control Review ("the 1996 PCR") whereas NTS calls were not.

3.4 The use of a baseline consistent with the 1996 PCR was chosen for the Oftel methodology used in the April 1999 and April 2000 directions rather than a starting point based on retail costs from BT’s current accounts (eg 1999/2000 accounts for the April 2000 direction). Oftel used the 1996 PCR as a baseline in order to avoid the possibility of BT over recovering its common retail costs by recovering them within the 1996 PCR and again in the retail uplift. This would happen because the enormous growth in NTS call volumes would, under BT’s FAC methodology, lead to a reallocation of costs towards NTS calls. Thus the same costs would be allocated to both PCR services in the 1996 allocation and to NTS services in the 1999/2000 allocation.

3.5 For the reasons set out above, Oftel rejected the broadbrush methodology in favour of its current methodology based on the explicit identification of relevant retail costs and the allocation of costs to NTS consistent with the 1996 PCR. An alternative starting point for the calculation of the retail uplift could be BT's most recent FAC numbers. One of the key features of Oftel's methodology is that the economies of scale effect due to past volume growth in NTS calls is reflected in the NTS retail uplift rather than being spread across all call types. By contrast, BT's FAC figures allocate costs across all call types according to volumes and therefore a retail uplift calculation based on these costs would lead to the economies of scale due to the growth in NTS calls being spread across all call types. The relevant EU market review (Review of the fixed narrowband wholesale exchange line, call origination, conveyance and transit markets, consultation - 17 March 2003) is considering a range of options for calculating the retail uplift in the future, one of which is an approach that uses BT’s FAC as a starting point. However, relevant FAC numbers are not available for the period of this direction and Oftel considers that its methodology used in the April 1999 and April 2000 Directions remains the best approach for the calculation of the retail uplift up to July 2003.

Details of the methodology

Baseline

3.6 As explained above, the starting point for the retail uplift cost allocation is the implicit allocation of retail costs to NTS services made by the 1996 PCR. In order to estimate the baseline figure for NTS retail costs, the unit retail cost figure is obtained by dividing the total retail costs (i.e. those shared between geographic and NTS services), by total volumes (Local, National and NTS Volumes) and multiplying this by BT’s NTS call volumes for the relevant year.

3.7 An allowance for the rate of return on capital employed (ROCE) is added onto the baseline figure for total relevant retail costs to allow BT to earn a reasonable return for providing retail services. (Article 7(2) of the Interconnection Directive 97/33 recognises that a reasonable rate of return is permitted in a cost oriented interconnection charge). This is BT’s real cost of capital multiplied by the mean capital employed of Local and National calls. This pro-rata methodology of devising the charge is underpinned by the assumption that the unit retail cost between geographic and non-geographic calls is the same.

3.8 Oftel considers this assumption to be reasonable, because Oftel believes that retailing activities are very similar between geographic and NTS call types. Oftel also believes that all such costs, or at least the vast majority of them, are therefore common between those call types. This is explained in greater detail in paragraphs 3.19 to 3.26.

The baseline calculation is summarised below:

3.9 A further adjustment is made to the baseline costs to reflect the fact that BT terminated NTS costs are found in the "Other" category of the RSB, not the Local and National call categories. Therefore, the total retail cost figure must be suitably increased to reflect these costs. This ensures that the cost and volume data are consistent. The actual baseline calculation is the following with the adjustment in italics:

Unit cost calculation =

The unit cost figure obtained from the above is multiplied by NTS volumes in 1994/5 to obtain the baseline allocation of costs. The adjustment is made using confidential BT terminated NTS volume data.

Methodology to update the allocation to the current year

3.10 As explained earlier, this initial baseline allocation is updated to reflect the NTS volume growth, inflation and efficiency savings between 1994/1995 and current year of the charge being directed. Each adjustment is explained in turn below.

Volume adjustment

3.11 Since the volume of calls has significantly increased it is reasonable to expect that costs are now higher (i.e. a positive Cost Volume Relationship ("CVR")). Oftel considers that if any element of cost is variable with the level of production, total costs rise with output. This would not be the case only if all costs were fixed costs.

3.12 Oftel does not believe that retail costs would increase on a one to one ratio with volume (eg so that a 100% increase in volume would lead to a 100% increase in costs, giving a CVR of 1). Oftel recognises that retail costs in the telecommunications sector exhibit substantial economies of scale, although the extent of scale economies varies with the precise cost being considered. For example, Oftel considers that costs such as marketing and sales (M&S) are unlikely to significantly increase with volumes, implying a very low CVR. Costs for activities such as customer services are likely to increase more in line with volumes, implying that a higher CVR could be expected.

3.13 In the 1996 PCR, Oftel took the view that an overall CVR of 0.25 reasonably reflected the average movement of retail costs in relation to volumes. Oftel believes that this continues to be a reasonable CVR for all costs except for bad debt which it treated slightly differently, as explained below. If a higher CVR had been used, the empirical data on costs and volumes would have implied implausibly high efficiency gains (these are dealt with further below). The CVR and the efficiency gains must both lie within a plausible range and must, together, reconcile with BT’s actual data on costs and volumes. A CVR of 0.25 satisfies these requirements.

3.14 Oftel considers that retail bad debt costs vary more closely with revenue (price), rather than volume. This is because the bad debt amount owed by a retail customer depends directly on the price of calls. The expected amount of bad debt is higher for an expensive call rather than a cheap call. Accordingly, Oftel considers that a Cost Revenue Relationship ("CRR"), rather than a CVR, should be applied to bad debt. Oftel further takes the view that the CRR is one (ie: costs of bad debt increase directly with revenues on a one to one ratio). This is because there are not likely to be any economies of scale associated with bad debts. Oftel applies the CRR to actual bad debts written off, which is 66.9% of BT’s Finance and Billing category. This figure has been derived from information provided by BT, during the time of the April 2000 Direction. BT provided a confidential statement giving revenue written off due to bad debt as a function of its total retail costs for local and national calls, including some NTS, for the period 1997/98 to 2000/01. This showed that bad debt formed, on average, 69.2% of BT’s finance and billing costs which equates to approximately 2.3% of turnover for these calls. Having reviewed BT’s submission Oftel considers that it is not unreasonable to rely on this evidence in calculating the final charge. In removing the PRS bad debt costs that had originally erroneously been included, Oftel has adjusted this figure to reflect this fact, and this figure is now 66.9%.  

Efficiency Adjustment

3.15 Oftel then makes adjustments for efficiency savings that are independent of volume changes. In other words, this recognises that BT’s Retail Systems Business ("RSB") has become more efficient since 1994/1995 so that retail costs have reduced. More generally, this is why one can observe that, although volumes of all call categories have increased over the time period, total retail costs have reduced, because the efficiency effect that reduces costs outweighs the volume and revenue induced cost increases.

3.16 Oftel calculated an annual efficiency factor for retail costs from retail cost and volume data for local and national calls. It did this by calculating the level of costs that would be expected in the current year as a result of volume increases (ie: by applying the cost volume relationship of 0.25 referred to above). Oftel then compared this estimate of total retail costs for local and national calls holding volumes constant. From these two figures Oftel derived an average annual reduction in costs due to improved efficiency. The formula behind this adjustment is presented in Annex 1. Oftel applied the efficiency savings factor to all retail costs except bad debt, because Oftel took the view that there is limited scope for additional efficiency savings in bad debts expensed.

3.17 Finally, Oftel adjusts the cost allocation with the Retail Price Index (RPI) to allow for inflation.

Composition of BT’s retail costs

3.18 The NTS retail uplift allows BT to recover a properly attributed portion of the relevant retail costs that it incurs, i.e. an allocation of costs that is cost oriented. BT has three main categories of direct retail costs. These are 1) customer service, 2) finance and billing and 3) marketing and sales. Oftel takes the view that customer services and finance and billing costs are directly relevant to NTS services, because BT incurs these costs whilst undertaking the retail billing functions of NTS calls. The Director therefore proposes to include these costs in the retail uplift charges.

3.19 Marketing and sales costs make up around 50% of the baseline retail costs. Oftel also takes the view that these costs are relevant to NTS services. These activities are treated as common between NTS calls and geographic calls, for the reasons given below. 

3.20 BT’s marketing and sales costs can be sub-divided into three categories:

• the first category captures marketing and sales activities that are aimed at increasing call revenue from existing customers;
• the second category comprises activities aimed at getting more people connected in the UK, either by connecting new customers or connecting customers who move from another operator to BT, sometimes referred to as "winback";
• the third category of marketing and sales activities relate to non-UK customers.

3.21 Marketing and sales costs which relate to non-UK customers (the third category) are not included in the retail uplift charge. Paragraph 5.5 of the April 2000 Direction stated that Oftel had included such costs into the final charge. This was an error. The relevant accounting categories from which the marketing and sales costs were extracted were the "local" and "national" call categories of BT’s Retail Systems Business. BT's cost attribution methodologies do not attribute marketing and sales costs related to non-UK customers into the local and national call categories.

3.22 In regard to marketing and sales costs attributable to the other two categories, Oftel takes the view that an appropriate portion of these costs should be included into the retail uplift charge. The reasons for this are set out in detail in Annex 1.

3.23 Having considered the matters described above, and which are described in detail in Annex 1, Oftel identified BT’s relevant retail costs (common between geographic and NTS calls) as:

• Finance and billing costs, 66.9% of which are due to retail bad debt, derived from an average figure over the period from 1997/98 to 2000/01, from information provided by BT – (see paragraph 3.14)
• Customer service
• Marketing – aimed at getting more people connected to BT in the UK
• Marketing – aimed at increasing call revenue
• Billing enquiries
• Fault report
• Complaints
• Indirect retail costs

Updating the retail uplift charge

3.24 In order to update the allocation of costs in the 1994/5 baseline to derive a retail uplift charge applicable from April 2001 to March 2002, and a further update from April 2002 to July 2003, Oftel has used BT originated NTS volume data provided by BT, which excludes all the volumes for unmetered services. Specifically, for the charge applied from April 2001, Oftel uses volume data for the financial year 2000/01. For the charge applied from April 2002, Oftel has used volume data for the financial year 2001/02. In addition, Oftel has updated the efficiency calculation with cost data from the relevant financial statements, and relevant years’ volume data from BT.

Chapter 4

Comments received to the draft proposals and Oftel’s responses

4.1 Comments were invited on the draft direction and were received from BT, Easynet, Cable and Wireless (C&W) and Energis. These can be found on Oftel’s website at:

www.oftel.gov.uk/publications/responses/2003/retup_0403/index.htm

The key points raised by each operator are summarised below with Oftel’s response given after each point.

BT

Process

4.2 BT commented that under Regulation 6 of the Interconnection Regulations, the Director may impose charges in a reference interconnect offer and "where an organisation makes changes to the published reference interconnection offer, adjustments required by the national regulatory authority may be retrospective in effect, from the date of introduction of the change". BT did not voluntarily make changes to the RIO with regard to NTS uplift charges but instead complied with Directions by Oftel to charge in accordance with a formula that Oftel adduced, and to which no alternative basis of charging was available.

Oftel’s response

4.3 The Director’s decision in a case such as this may be retrospective in effect, as Article 7 of the Interconnection Directive makes clear. This is so, whether changes to an interconnection contract are made voluntarily or as a result of the exercise by the Director of his functions under the Directive.

Methodological validity

4.4 BT said Oftel had stated that consistency with the 1996 PCR is important in terms of cost recovery. This was because a certain amount of retail costs were assumed to be recovered from NTS calls, and hence excluded from consideration for the retail call price control. Oftel also stated that "had volume forecasts been growth been anticipated, NTS retail costs would have been capped."

4.5 BT added that Oftel’s treatment of the retail uplift is inconsistent with the PCR, because no retail price cap has ever been re-opened due to volume growth deviating from that assumed by Oftel at the outset. Oftel’s intervention is equivalent to re-opening a price cap, not just because unexpected events have occurred (NTS volume growth), but because it has itself allowed the effects to accumulate.

Oftel’s response

4.6 NTS calls were not included in the 1996 PCR. Thus Oftel’s adjustment to the retail uplift is not reopening that control. Neither is it equivalent to re-opening the price cap since the level of the retail control was not dependent on any specific volume forecast for NTS calls.

4.7 The growth of NTS is largely due to internet calls, which are incremental to geographic calls included within the price cap. It appears implausible that the such a large growth in NTS call volumes is substitutional to geographic calls. The growth of NTS volumes necessitates the regulation of the retail uplift in a way that accounts for economies of scale generated by this volume growth. Oftel believes that the approach taken appropriately distributes benefits to customers making use of non geographic consumers calls (rather than consumers making geographic calls).

Efficiency gains

4.8 BT suggested that the NTS volume growth might warrant breaking from the Price Cap methodology and making a correction for forecasting errors. However, the approach to achieved efficiency gains has no justification. For consistency with the PCR, the same forecast efficiency gains should be used in the methodology.

4.9 BT argued that it may have out-performed the efficiency target set in 1996 and is benefiting as a result. Therefore, the proposed methodology would simply deprive BT of these gains as they apply to one type of call (NTS) and would transfer these cost reductions to terminating operators. As such as system is not part of any price control, nor has it ever been, this also represents a deviation from the methodology underpinning the PCR.

Oftel’s response

4.10 BT are implying that Oftel should have used forecasted efficiency gains rather than out-turns because the use of out-turns undermines their incentives for cost reduction. Oftel take the view that this may be true if there was a price cap in force, but this is of course not the case for NTS calls.

Use of 1996 ‘CVR’

4.11 BT claimed that the CVR used by Oftel has no substantive basis. Furthermore, Oftel has made the incorrect assumption that all of BT’s retail costs have a linear relationship with volumes.

4.12 BT considers that the MICRA data supplied to Oftel provides a more appropriate CVR than that used by Oftel. If used, a MICRA based CVR would produce a charge which is the regulatory cost floor of [confidential]. This is higher than Oftel’s proposed charge of 0.1954ppm.

4.13 BT, in referring to paragraph 8 in chapter 5 of the Draft Direction, claimed that a Marketing & Sales CVR of 0.05 is unsubstantiated.

Oftel’s response

4.14 Oftel based the CVR on assumptions that were used in the PCR. Oftel believes that this was a more consistent approach compared to using the MICRA data, since the latter has produced, in Oftel’s view, an implausibly high CVR, given the nature of retail costs.

4.15 In practice, retail costs have fallen over time against increasing volumes and this can only be explained by the efficiency savings made by BT. If a CVR of the magnitude suggested by BT is applied to the calculation, then the efficiency savings factor resulting from the above formula would be even greater, around 9% in order to be consistent with the observed reduction in costs over time on application of the above formula.

4.16 Moreover, Oftel believes that BT’s retail CVR estimate is not credible, and does not believe that the LRIC/FAC ratio is the appropriate way to determine a retail CVR.

4.17 In paragraph 5.8 of the draft direction Oftel gave an example to illustrate the impact of CVRs on the retail uplift calculation. The figure of 0.05 was stated to be an assumption within the example. This is not a definitive view on the specific CVR for M&S, the retail uplift calculation uses the assumption that M&S costs are likely to have a very low CVR.

Recovery of costs

4.18 BT pointed out that an amount of costs allocated in BT’s accounts have been excluded from the RU charge, but have not been re-allocated to other calls. BT considers this is absurd.

4.19 BT claimed it is under recovering its costs. BT added that Oftel may argue BT has allowed an RPI minus RPI cap on geographic calls to take this into account. If this is the case then Oftel should have taken this into account in the latest PCR negotiations. Even if this is true this does not account for costs before 1^st August 2002. The proposed Draft Direction represents a subsidy of OLO terminated NTS calls by BT geographic calls. 

Oftel’s response

4.20 Oftel has allowed BT to recover costs that are actually incurred in providing a service, thereby ensuring its charge is cost orientated.

4.21 Prior to the 2002 price control review, an up to date FAC approach for the setting of the retail uplift was not considered appropriate because of the existence of the 1996 price control review (PCR). It is important that other retail regulation such as the retail uplift is consistent with the retail cost allocations underlying the 1996 PCR. The use of up to date FAC costs would have been inconsistent with the PCR that was in force at the time, because up to date cost allocation allows a different allocation of costs between geographic and non-geographic calls. This would have had the effect of BT over-recovering in retail costs, once through the price control review, and then again through the NTS retail uplift.

4.22 However, the situation is different now, given the 2002 price control review. The problem of over recovery through the use of up to date FAC data no longer precludes the use of more up to date data, because the 2002 price control review is based on up to date FAC data. However, the use of the current methodology (1994/5 baseline) is also not inconsistent, because the 2002 retail price cap is not expected to be the primary constraint on call prices. The main constraint on BT’s retail prices is expected to be provided by competition. Accordingly the costs which BT can recover are governed by competition rather than by the allocation of costs between geographic and non geographic calls. This implies that there are no significant problems of inconsistency in using either approach; i.e. the previous allocation underlying the 1996 PCR, or the up to date cost allocation. As pointed out in paragraph 3.5, Oftel’s methodology has the feature that it captures, for customers of NTS calls, the economies of scale associated with the growth in NTS calls.

Incorrect 2002/03 charge

4.23 BT pointed out that there was likely to be an error in this charge, because the reduction in volumes, and increase in costs between this year and the year before indicate that the charge should go up.

Oftel’s response

4.24 Oftel has checked the calculations and agrees with BT on this point. A data input error has been found and corrected. The April 2002 charge is now higher than that for the previous year.  

Marketing and sales costs

4.25 BT stated that the percentage of NTS retail costs represented by marketing and sales is 27% as opposed to 50% as implied by Oftel. This indicates that Oftel has weighted the retail uplift charge too heavily towards marketing and sales.

Oftel’s response

4.26 Oftel made the point in the draft direction that a low overall CVR of 0.25 implies that the CVR accruing to marketing and sales is likely to be very low, implying that the charge has a small weighting of M&S.

4.27 The 50% figure used by Oftel is the upper limit for the proportion of M&S costs, i.e. it includes direct costs that support M&S, such as accommodation and computer equipment. It is therefore not inconsistent with BT’s figure which is the lower limit consisting of only direct costs. The lower limit assumed in the model is 28%, which is virtually the same as BT’s figure of 27%. This implies consistency with BT’s cost breakdown.

Other operators

Freephone calls

4.28 Easynet argued that the inclusion of freephone calls (excluding bad debt costs) should reduce the overall retail uplift because the inclusion of more freephone minutes means that the overall level of costs should have reduced (on the basis that freephone minutes cost less to bill than charged for minutes).

Oftel’s response

4.29 Oftel has not disaggregated the methodology to the level of detail of attributing costs by each different call type. Oftel does not take the view that freephone calls attract less retail costs, other than bad debt, than other NTS call types. Oftel did remove bad debt costs from the freephone retail uplift. This was because this adjustment was likely to have a material impact.

Variation in the retail uplift from year to year

4.30 Easynet made several points about the methodology based on the fact that the retail uplift for 2001/02 was much higher than the charges for the previous and following years.

Oftel’s response:

4.31 Since the Draft Direction, in response to points highlighted by BT’s consultation response, Oftel has reviewed the calculation and found some data input errors in the local and national volume data used in the efficiency calculation, which have the effect of changing the relative differences between the retail uplift charges. The new charges are summarised below in Table 1. For ease of reference, table 2 gives the charges that Oftel consulted in the Draft Direction, and Table 3 gives the volume figures for the relevant years:

Table 1: Retail uplift charges in Final Direction

Table 2 :Retail uplift charge proposals in Draft Direction

Table 3: BT originated NTS metered volumes

4.32 The average retail uplift figure increases in the final year, and the average charge from April 2001 decreases slightly (table 1). This can be explained by the decrease in volumes between 2000/01 and 2001/02.

4.33 In order to understand the movements in the relative charges, it is useful to consider the relationship between costs and volumes. In general, the growth in unit costs ("UC") is a function of growth in total costs ("TC") and growth in volumes ("V").

4.34 The impact of volume growth on total costs is complicated but the main points are:

• falls in volumes reduce total costs due to the link of volumes and costs via the cost volume relationship; and
• falls in volume (and price) reduce revenues, which are fundamental to the fall in bad debt; but
• a fall in volume can also increase unit costs, if there are now less volumes over which to spread a fixed amount of costs.

4.35 Given the fact that lagged data is used in setting the retail uplift charge, the total cost, unit cost and volume data refer to the previous year to which the charge actually applies. Consider the total costs ("TC") between 2001/2 and 2000/1.

Total costs for the charge for 2000/01 was £117,063,357 and that for 2001/02 is £98,141,021. The ratio between them is:

98,141,021/117,063,357 = 0.838 (1)

The ratio between volumes is

48017/61258 = 0.784 (2)

This implies that total costs have not fallen as much as volumes, which indicates that the unit cost should increase.

The ratio of the unit costs is

0.2044/0.1911 = 1.070 (3)

4.36 This figure is exactly equivalent to the ratio of (1) and (2), i.e. the change in total costs divided by the change in volumes = 0.838/0.784. This implies that the figures are internally consistent, and that the behaviour in total costs and volumes explains the behaviour in unit costs.

Lagged data

4.37 C&W stated that Oftel should use actual data for the relevant year instead of data from the previous year. C&W maintain that any deviation from applying actual cost data to build BT’s charges will afford BT the opportunity to over-recover costs as a whole.

Oftel’s response

4.38 In general, Oftel uses lagged data in setting the retail uplift charge, because current data is not normally available, and lagged data is often a better proxy than forecasts. However, in this case, due to unforeseen circumstances there was a series of delays which led to these charges becoming retrospective.

4.39 However, despite this, Oftel believe that it is reasonable to continue to use the lagged data approach, because this allows Oftel to maintain consistency by undertaking an approach that would have been undertaken in the absence of the delay, so there are no systematic gainers or losers from delay.

4.40 Moreover, given the fact that, in the light of the consultation, the retail uplift charge figures have changed since the Draft Direction, it is clear to Oftel that BT do not significantly over recover (or indeed under recover) in using the lagged data approach.

Retrospection

4.41 C&W stated that it found Oftel’s decision to retrospectively apply the retail uplift charges inconsistent with Oftel’s decision not apply the bad debt surcharge decision retrospectively (Direction under the provisions of Regulation 6(6) of the Telecommunications (Interconnection) Regulations 1997 resolving a dispute between Cable & Wireless Communications (Mercury) Limited ("C&W") and British Telecommunications plc ("BT") over a bad debt surcharge relating to calls to Premium Rate Services).

Oftel’s response

4.42 In the case of the retail uplift, the original April 1999 direction was a retrospective decision because of the exceptional circumstances given in that direction. In brief these were:

• BT had failed to review the retail uplift after August 1998; and
• operators had a legitimate expectation that the retail uplift would be reviewed regularly following statements made by the Director:

4.43 The amendments to the April 1999 and April 2000 directions do not represent retrospective decisions but are correcting an error discovered after those decisions were made. This direction relates to the period 1 April 2001 to 24 July 2003 for the reasons given in paragraph 2.6. These are namely that Oftel’s initial intention to set a price cap from 1 April 2001 had to be put on hold pending Energis’ appeal of the April 2000 direction. This was followed by the announcement of the market reviews required by the new EU Directives. These events meant that the initial timing of the price cap became inappropriate and will now be considered within the context of the market reviews. Oftel had therefore to set charges to cover the intervening periods.

4.44 The PRS bad debt decision is not retrospective because of the absence of any exceptional circumstances which would justify such a decision. The date of the direction is the date on which C&W issued BT with a price change notice. This is consistent with Oftel’s resolution of other disputes.

ROCE and efficiency

4.45 Easynet questioned the level of ROCE which is added to the baseline figures and argues that the inclusion of a ROCE in BT’s baseline figures does not provide BT with the correct efficiency incentives. Easynet also questioned Oftel’s use of the ‘efficiency adjustment’ and states that Oftel should only allow BT to recover costs incurred by an efficient operator and OLOs should not be paying for BT’s inefficiencies.

Oftel’s response

4.46 Cost orientation permits the inclusion of a reasonable rate of return on investment. In accounting terms this is assessed as a reasonable rate of return on the capital employed by the firm ("ROCE"). ROCE is a measure of the accounting profit earned by firms (usually before interest and tax). It is expressed as proportion of a firm’s investment in the business. ROCE is calculated by dividing the earnings before interest and tax by the mean capital employed. In a competitive market, the ROCE would be expected to equate roughly to the cost of capital which is the return on capital needed to attract investment to a firm. In the calculation of a regulated charge, it is appropriate to include an allowance for a reasonable rate of return (ie: the profit that would be expected in a competitive market). Without an allowance for the cost of capital, the regulated firm would not undertake the investment necessary for it to continue to supply the service.

4.47 Oftel disagrees with Easynet’s comment on the efficiency adjustment. Easynet have not been paying for BT’s inefficiencies, because the methodology takes account of the efficiency improvements due to competition and technical enhancements that BT has achieved from 1994/5 to the current year. Oftel have captured these effects to allow OLOs to benefit from several years of efficiency gains.

Bad debt

4.48 Easynet asks why BT’s bad debt increased from 66.9% to 69.2%.

Oftel’s response

4.49 BT’s bad debt proportion of Finance and Billing costs, in fact, fell from 69.2% to 66.9%, because of the revelation that the former figure contained excess PRS bad debt. When this was removed, the percentage of bad debt associated with NTS calls reduced accordingly.

Marketing and sales costs

4.50 Easynet said it believed that the cost of BT’s ‘winback’ activities and the cost of marketing to new connections should be excluded. Even though Oftel states that, because of the very low CVR applied to sales and marketing costs, the impact on the retail uplift is very small, Easynet believe that Oftel should exclude these costs. 

Oftel’s response

4.51 Oftel has not excluded these costs for two main reasons. First, it is difficult to determine whether such costs are irrelevant to NTS calls (by being incremental to geographic calls). In the absence of robust evidence indicating that such M&S campaigns were incremental to geographic calls, Oftel has taken the view that all M&S costs were common between NTS and geographic calls.

4.52 Second, the costs included in the calculation of the retail uplift are ultimately paid by customers. The test as to the relevance of particular cost items is therefore whether it is reasonable for customers to pay towards these costs, not whether an operator such as Easynet benefits from those costs being incurred. In Oftel’s view, it is reasonable for customers to pay some amount towards BT’s general costs of attracting and maintaining its subscriber base in the prices paid for NTS calls.

4.53 This is because marketing and sales costs may be expected to increase volumes of subscribers and calls going over the network. These increases in volumes lead to reductions in unit costs through economies of scale. Such reductions in unit costs ultimately benefit all customers because, as costs per customer are reduced, prices to customers can be reduced. To exclude the marketing and sales expenditure would be an inconsistent and one sided calculation: consumers would benefit from the reductions in costs (driven by increased volumes) but would not be required to pay for the expenditure that led to the reduction in cost in the first place. Oftel has taken this view in setting charges for geographic calls, and considers it equally applicable to NTS calls.

Implementation of directions

4.54 Oftel has noted operators’ concerns over its initial proposals which allowed BT to implement Oftel directions without having to seek operators’ agreement through having to sign pricing letters. As a result Oftel’s proposals have been modified to reflect as near as possible the processes contained in paragraphs 13.3 to 13.7 of the BT Standard Interconnect Agreement (SIA). The key difference being that operators will no longer be required to signify, in writing, their acceptance of BT’s change resulting from the direction. Details of the revised process can be found at paragraphs 5.11 to 5.17 of this document.

Chapter 5

The Director’s final decision

The relevant market and the position of BT

5.1 BT’s retail uplift forms part of the price for an interconnection product. For the purposes of the Interconnection Directive (97/33) (the ICD), BT has been determined as having Significant Market Power (SMP) in the markets for fixed public telephone networks and services, and is therefore required to offer interconnection to operators with Annex II status. Article 7(2) of the ICD requires BT to show that its charges are cost oriented and for the national regulatory authority (i.e. the Director) to require that charges be amended where they are not cost oriented. The Director has a general responsibility in Article 9(1) of the ICD to encourage and secure adequate interconnection in the interests of all users, exercising his responsibility in a way that provides maximum economic efficiency and gives the maximum benefit to end-users.

5.2 The Director’s final decision is made in accordance with Article 9(3) of the ICD (as implemented in Regulation 6(3) of the Telecommunications (Interconnection) Regulations 1997 in pursuit of these aims taking into account the matters listed there (as implemented in Regulation 6(1)).

Exceptional circumstances

5.3 Article 9(3) of the ICD and Regulation 6(3) of the Regulations state that "The Director may intervene at any time and that he may in exceptional circumstances make a direction that changes be made to interconnection agreements already concluded where it is justified to ensure effective competition or interoperability of services for users or both."

5.4 The exceptional circumstances justifying this final direction are:

• a history of long running disputes about BT’s retail uplift;
• the need to ensure cost orientation and economic efficiency by adoption of a specific Oftel methodology to meet these aims in order to calculate the retail uplift year by year;
• the legitimate expectation of the industry that Oftel would continue to control BT’s retail uplift from April 2001; and
• the interaction and timing of the resolution of the disputes and this own initiative investigation with the EU market reviews and the change in regime in July 2003.

The Director’s final decision

5.5 The Director’s final decision is that, with effect from 1 April 2001 to 24 July 2003 inclusive, the methodology for calculating BT’s NTS Retail Uplift shall be as described in Chapter 3 of this explanatory memorandum and the revised charges having used that methodology shall be:

An average retail uplift charge of 0.1911ppm, which gives rise to:

An average retail uplift charge of 0.2044ppm, which gives rise to:

5.6 It was Oftel’s original intention to propose the retail uplift charge for the year 1 April 2001 to 31 March 2002, the year 1 April 2002 to 31 March 2003 and for the period 1 April 2003 to 24 July 2003. The latter period is a shortened period recognising that a new legal regime will enter into force on 25 July 2003. What the obligations will be under that regime from 25 July 2003 are currently the subject of market reviews which are being carried out as preparatory work to that new regime.

5.7 However, Oftel has had difficulty in attempting to forecast with any certainty the volumes in 2003 so as to calculate a retail uplift charge for 1 April 2003 to 24 July 2003 based on those forecasts. This is because, being forward looking, there is no reliable way of forecasting the unknown extent that calls to metered NTS internet services will migrate to unmetered and broadband services in the year 2003/2004. The alternative option of waiting for the relevant data to become available is likely to be impractical because a new legal regime will have entered into play before that data is available. Oftel therefore proposes that the retail uplift charge derived for the year 1 April 2002 to 31 March 2003 should be extended to apply until 24 July 2003. In light of the new regime entering into force on 25 July 2003, the desire to move away from dealing with issues on a backward looking basis, and to move towards certainty, Oftel considers that this is a reasonable proposal in these circumstances. 

Implementation of the Direction

5.8 In the draft direction, Oftel proposed changes to the process used by BT to implement Oftel directions through publication of pricing letters and OCCNs.

5.9 A number of operators expressed concerns about how, under the proposed new process, perceived errors in BT’s implementation of the Director’s decision would be addressed. Oftel has considered these concerns and concludes that the implementation process in this direction should offer similar opportunities in terms of time for highlighting genuine errors in BT’s change proposals and subsequent negotiation and dispute, as the process already in operation for OCCNs.

5.10 The main difference compared to current OCCN procedure, for the purposes of implementing this direction, is that the Operators will not have to notify their approval of BT’s proposed changes within 14 days of receipt of the proposed changes. Instead, such approval will be deemed after 14 days in the absence of the Operator notifying BT otherwise as set out below.

5.11 Accordingly, and only in relation to the changes resulting from this direction, as soon as possible following publication of this final direction, BT will notify operators of the changes it proposes to make to the CPL in order to implement this direction. Operators will then have 14 days from receipt of the proposed changes to notify BT in writing of any concerns with BT’s proposed changes to the CPL stating why that Operator believes that the proposed changes do not correctly implement this direction. If an Operator is satisfied that the proposed changes do correctly implement this direction, there will be no requirement to notify BT of its approval of the changes.

5.12 If, after the 14 days, no formal concerns are notified to BT, the proposed changes to the CPL will be deemed to have been approved and BT shall include the changes in the CPL.

5.13 Where an operator does raise concerns that BT’s proposed changes do not correctly implement this direction BT and the operator shall have a further 14 days to seek to reach agreement as to the proposed changes so that they do correctly implement this direction. If the parties fail to reach agreement within the further 14 days either party may, not later than 1 month after the expiry of the second 14 day period, refer the matters in dispute to the Director. This process provides a means of bringing a dispute on the incorrect implementation of this direction and is not a route that can be used to reopen the issues already dealt with by the Director.

5.14 Until agreement of the proposed changes is reached by the parties, or the proposed changes are determined by the Director as a result of a dispute, the charges currently contained in the CPL shall continue to apply. The new changes, once agreed, shall then be included in the CPL.

Annex 1

Why Marketing and Sales costs should be included in the Retail Uplift

1. Oftel initially sought to identify which marketing and sales costs were incremental to particular services. The test for whether a particular element of cost is incremental to a service is to consider the difference in costs between the situation in which the service is provided and the situation in which it is not provided. This difference is the cost which is incremental to that particular service. Costs which are not incremental to any particular service must logically be regarded as common among a number of services.

2. Oftel has previously sought information from BT from which it could be established that marketing and sales costs should be included into the retail uplift charge. BT’s response stated that it was not possible to provide conclusive evidence of the kind that Oftel had requested. BT submitted that the only way to prove such causality would be to run a controlled test under precisely the same conditions but without running the campaign, and then to observe the differences in volumes. BT said that marketing campaigns do not lend themselves to such an approach.

3. Oftel was persuaded by BT’s response as to the impracticality of establishing causality in this case. The precise identification of all of BT’s marketing and sales campaigns would not be sufficient in itself to enable an assessment of causality to be made. In addition, a controlled test would need to be carried out so that an assessment could be made of the impact of the campaign on call volumes and call type. Oftel considers that it would be impracticable and disproportionate for BT to carry out such controlled tests, particularly since (as explained in the next paragraphs) it was not expected that such tests would establish that a substantial part of such costs were other than common between geographic and NTS calls.

4. Oftel has considered the various categories of marketing and sales costs and considered that all, or at least the majority, are common to both geographic and NTS. BT has submitted to Oftel that its marketing and sales costs were divided between price-related campaigns and general call-stimulation campaigns. Oftel was satisfied that marketing campaigns like "Friends and Family" are common to both geographic and NTS calls. Under these campaigns, BT’s subscribers may, for example, include NTS calls or geographic calls as their "Best Friend" telephone numbers and hence receive a discount on calls to those numbers. General call-stimulation campaigns are aimed at the stimulation of all calls, whether geographic or NTS, and whether voice or data calls. 

5. It was against this background that Oftel has had to take a view on the inclusion of marketing and sales costs into the retail uplift charges.

6. Oftel proposes to include a reasonable portion of all retail costs in the NTS retail uplift (ie: to treat them all as common for the reasons set out above), rather than attempt to assess in detail which costs (if any) were incremental to which services. This is because there is no evidence available from 1994/5 as to which costs if any are likely to be incremental to particular services. Even if this information on 1994/5 costs was available now, there is no evidence to suggest that it is relevant to the current year. Accordingly, Oftel has treated all retail costs at the 1994/5 baseline as common and allocated a portion of those common costs to NTS. Oftel is not aware of any marketing and sales campaigns which are incremental only to geographic calls.

7. Marketing and sales costs feature only very slightly in the final charges in this direction. This is because, even though marketing and sales costs are included in the baseline, Oftel has applied a very low CVR to them (ie: a weighted average of 0.25). As explained in paragraph 3.12, Oftel considers that marketing and sales costs are more similar to those of a fixed nature, and therefore do not significantly increase in line with volume. This has the effect of significantly decreasing their proportion of the retail costs alongside the volume induced increase in retail costs. This can be explained through a simple numerical example using dummy numbers.

8. Consider a 1994/1995 baseline cost allocation of £100, with an initial volume of 10 in 1994/1995. Assume that 50% of these costs are marketing and sales costs. Assume marketing and sales has a CVR of 0.05, and this is reflected in the overall CVR of 0.25, which is applied to the entire allocation of costs. Assume volume has grown from 10 to 600 over 6 years. This is an implied growth of 2900%. After 6 years, the costs will have grown by:

0.25*2900% = 725%

The cost at the end of the 6 years will be: