Downloaded 16 Jan 2002

Mr Paul James
Regulatory Policy Directorate
Oftel
50 Ludgate Hill
London EC4M 7JJ
14 August 2000
Dear PaulOftel's draft Direction against BT Cellnet for unfair discrimination
The Service Provider Interest Group, SPIG, which represents the interests of independent communications service providers, wishes to endorse and support the comprehensive response by the FCS Independent Mobile Service Providers Group, to the Draft Direction under Condition 66 of the Mobile PTO Licence granted to Telecom Securicor Cellular Ltd issued on 3 July 2000.
In particular:
Our overriding concern is that customers will be the losers if unfair market conditions such as cross subsidies continue, which keep prices high and stifle competition. We support Oftel in putting in place this Determination and look for vigorous action in future.
Yours sincerely
Phil Sayer
Chairman, SPIG
This paper forms the response of the Independent Mobile Service Provider Group to the OFTEL draft direction issued on 3rd July against Telecom Securicor Cellular Limited ("BT Cellnet").
The Independent Mobile Service Provider Group, the Group, is an autonomous Group within the Federation of Communication Services representing the majority of the active Independent Service Providers competing in the mobile market in the UK. Therefore this response is made on behalf of those who will be most directly affected by the eventual OFTEL decision.
The Group takes an active role in representing the interests of our members at OFTEL Working Groups and other industry forums. Together members of the Group have unparalleled experience of the commercial realities of the UK mobile market. This response is written with the full support of all of these members.
The Group has already made clear its extreme dissatisfaction and disappointment at OFTEL’s decision, without reference or explanation to any independent service providers, to limit the investigation to the first Quarter of 2000 and to disregard the clear evidence of BT Cellnet and Vodafone failure to comply with the OFTEL Formula during 1997-1999 (which was set out in OFTEL’s Statement published in February 1999). That decision has been compounded by the length of time which OFTEL has taken to pursue this investigation, which commenced in July 1999. In the Group’s view these deficiencies directly contributed to the acquisition of many Independent Service Providers by BT Cellnet and Vodafone during that period.
The Group reluctantly accepts that it must now pursue those concerns separately and outside the framework of this consultation process. However, the Group believes it is only right to put those overriding concerns on the record so this response cannot be misread as an overall endorsement of OFTEL’s approach.
The remainder of this response, therefore, concentrates solely on the detail of the OFTEL investigation into cross-subsidies between BT Cellnet and Vodafone in Quarter 1 of this year and on the proposed action against BT Cellnet for unfair cross subsidy during that period.
The Group supports the proposed Direction against BT Cellnet and Independent Service Providers will be looking for an appropriate reduction in the subscription and prepaid price. However, the Group believes that OFTEL has underestimated the extent and seriousness of cross-subsidy by both Vodafone and BT Cellnet during the period under investigation.
OFTEL concludes1 that average subscriber life for subscription services is 22 to 26 months and for pre-pay is 45-56 months. The Group accepts the underlying methodology2 but considers that these figures considerably overestimate average subscriber life for pre-pay services. The difference stems from OFTEL’s determination of churn rates for that sector of the market.
OFTEL has concluded that monthly churn for pre-pay services is below 2%. However, this considerably underestimates the true churn figures for the following reasons:
The Group’s conclusions are that these corrections to the OFTEL assumptions will take pre-paid churn to similar proportions to churn in the subscription sector of the market and produce a subscriber life of at (or even below) 22 months rather than OFTEL’s conclusion of 48-52 months7. This will of course, have a corresponding effect on the point at which Service Providers can recover their costs. It will inevitably uncover the cross-subsidy that has been taking place but which was not apparent once OFTEL had arrived at artificially inflated periods for pre-pay subscriber life. With the vast majority of UK subscribers opting for pre-paid services this will have a dramatic effect on OFTEL’s calculations.
This is the other component of the OFTEL Formula. Here, by contrast, members of the Group broadly accept OFTEL’s assumption of 9.26% per annum as the pre-tax real cost of capital.
The generality of members of the Group have found it impossible in practice to take advantage of the availability of wholesale pre-pay connections because margins (at or even below cost) have been too low to make the business viable. In practice only Tied Service Providers have been able to absorb those margins because of hidden cross-subsidies from the parent networks. Therefore the following examples concern margins at the distribution level (which affect many members of the Group) as well as wholesale pre-pay connections. Together they illustrate the hidden cross-subsidies which the Group considers OFTEL must investigate fully before coming to any conclusion on the relationship between Tied Service Providers and their parent companies.
According to published accounts14 the following Vodafone Tied Service Providers have been trading at a loss:
The Group is concerned that these accounts could constitute further material evidence that Vodafone Tied Service Providers have been receiving unfair cross-subsidies from their parent company. (Unprofitable margins) "may be viable in the context of the overall Group within which a TSP operates because the parent company is likely to be generating profits through supply of wholesale services [but] it would not enable ISPs to earn a reasonable return15".
It would simply not be viable for members of the Group to run their companies at a loss. The Group asks OFTEL to include these concerns in its investigations.
The Group is extremely concerned that OFTEL was unwilling to publish the data which it has received from BT Cellnet and Vodafone in response to the latest investigation. This is in sharp (and unexplained) contrast to previous policy in 1999. Then OFTEL published the number of months each Tied Service Provider would require to cover acquisition costs and achieve a reasonable return16. The Group has not been given a satisfactory reason for this change in policy17 - which has made it much more difficult to comment on the investigation than should have been the case.
The Group is, through this response, now formally requesting publication of these details as soon as possible in line with previous OFTEL policy.
The Group supports continuing use of the Formula as the yardstick against which unfair cross-subsidy is measured. However, OFTEL must put a structured monitoring programme in place to ensure the parameters of the Formula are kept up to date. This current unsatisfactory situation is a direct result of OFTEL's longstanding failure to take any action against network operators abuse of their obligations under the OFTEL formula. Otherwise the market will simply move on again and OFTEL will be left to "play catchup" - unable to enforce the Formula because it is uncertain whether it should be updated. Competition and innovation will once again be damaged. The Group would like to work with OFTEL in achieving this and will be happy to supply relevant data on a regular basis. After such a long-standing failure to take action against the network operators abuse of the market it is the Groups belief that OFTEL must publish the quarterly returns and must commit to take action if network operators fail to meet their obligations
The Group believes that the only effective remedy would be a reduction in wholesale prices both for the subscription and pre-pay sectors of the market. The alternative - a decrease in the retail price would be disastrous for Independent Service Providers because it would effectively intensify the margin squeeze by other means. Only Tied Service Providers could then afford (again because of hidden cross-subsidies) to compete in the market.
Further the Group asks OFTEL to note that any reduction in the wholesale price should be applied proportionately to the amount of cross subsidy to each of the subscription and pre pay service provider tariffs. To apply any wholesale price reduction to pre pay only would largely be of immediate benefit to Tied Service Providers18 rather than the generality of independent service providers. Indeed Independent Service Providers have not been able to introduce pre-paid propositions based upon the wholesale purchasing of airtime as it has not been commercially viable. The network operators must not now benefit from their failure to offer a viable wholesale proposition
FOOTNOTES
1 Paragraph 13 of the Consultation back
2 Paragraphs 11-12 back
3 Only two service providers take wholesale pre-pay services from the Vodafone network: Vodafone Ltd (The Direct Business) and only one Independent Service Provider - which is a member of the Group. At the corresponding level on the BT Cellnet network only the Direct Business takes wholesale pre-pay services. The absence of Independent Service Providers is in large part explained by the extremely low margins (dealt with in more detail below) back
4 Because of the delay in recognising churn in pre-pay the commercial opportunities to profit with this behaviour are much greater than with subscription phones back
5 Therefore within the time frame of this investigation. back
6 The Group's research shows that of the new customer "Pay as you Talk" all calls base 25.8% of customers were inactive. Confidential information. back
7 This, could even be an underestimate. Experience shows that churn in the first six months following connection can exceed 8% per month. back
8 Confidential information. back
9 Since this attributes marketing costs directly to their objective - to win new customers. back
10 The Group would like to emphasise that the networks solely incur those costs to win new customers for their own brand (via Tied Service Providers) - not for generic reasons of building up the mobile market. Nor are the figures negligible - one member of the Group estimates they could be in the region of £100 per customer. back
11 Which are a self-perpetuating form of cross-subsidy - as Vodafone Tied Service Providers take further market share from Independent Service Providers which are unable to compete in the market, so the central Vodafone purchasing function has greater bargaining power with the mobile phone manufacturers. It can negotiate bigger discounts based on increasing market share and thus can ensure that manufacturing capacity is not available to satisfy orders from other sources. back
12 Again the Group can supply further details to OFTEL if required. back
13 Paragraph 17 back
14 Extract from Vodafone Annual Report, 31 March 2000 attached back
15 Chapter 3 of the February 1999 statement. back
16 For example in the Febrary 1999 statement at Table 3.1 back
17 Other than of potential unfairness to Vodafone as no licence breach was found for the period under investigation. However these reservations would apply equally to BT Cellnet and Vodafone for the previous investigation because OFTEL eventually failed to determine that either had committed a licence breach in that period. back
18 The one Group member that has been actively pursuing wholesale pre-pay connections would also benefit but other members of the Group would not. back
The argument on pre-pay churn (the figures used here relate exclusively to Vodafone)
"11. As a first step to estimating the average life of a subscriber OFTEL has sought to confirm whether the appropriate pattern of decay of the entire subscriber population continues to be inversely exponential (i.e. a function of the form y=Ae-kx which decreases at a decreasing rate and tends to zero over time). On the basis of survey data gathered in 1999, OFTEL is satisfied that this continues to be the appropriate decay pattern. In other words, in any given group of customers who purchase mobile telephony services at the same time, some customers will churn relatively quickly, while the probability that others will stay connected for a longer time increases as time goes by."
This is new, in the past it has been assumed that churn was a constant amount (e. g. 2% of the remaining base; period 1, 2% of 100%; period 2, 2% of 98%; period 3, 2% of 96.04% etc.)
Without knowing the values of the constants "A" and "k" used by OFTEL, which would have been most helpful, we agree that this method is reasonable given that there is no contract with the customer that includes an initial term. It is a characteristic of post pay contract customers that there is default churn within the first three months of the contract and excess churn in the three months following the expiry of the initial term when many customers are persuaded to churn. This is not the case with pre pay where there is no initial term and the contract is only as good as the customer buying the next top-up.
The following chart of an inverse exponential curve which no more than illustrates that the probability of churn is higher in the period following connection.

| 01/04/98 | 30/09/98 | 31/03/99 | 30/09/99 | 31/03/00 | |
| Net connections 12mths ended | 1648000 | 3233000 | |||
| Net connections 6mths ended | 1345000 | 1988000 | |||
| Customer base Source Vodafone |
198000 | 419000 | 1846000 | 3190000 | 5079000 |

In the six month period ended 31/3/2000 the subscriber base had a net growth of 1988000 or 59.2%. Vodafone does not recognise that a customer has churned until the expiry of six months following the last use. Therefore there is six months of churn which is unrecognised and this churn is predominately on the customers connected during the period, in other words on the average (approximately) of the customers in the period that is (5079000+3190000) ÷ 2 = 4135296.
In accordance with the OFTEL decay curve for subscribers, see (1) above, we can reasonably expect the probability of churn amongst these subscribers to be highest immediately following activation (this does not include those handsets that never actually connect). Experience shows that churn in the first six months following connection may be as high as 40% (on one sample, 2500 out of 6000 connections).
Accepting this assertion, if 40% of subscribers churn in the first 6 months and, these are not counted by the operators, the net growth figure for the period does not take account of the abnormal desertions which have not yet been recognised. This might be treated in one of two ways as far as the OFTEL formula is concerned, either:
In reality, therefore, the figures which the networks have provided to OFTEL do not measure pre pay acquisition costs and earnings for the same time period. Rather they measure earnings today but acquisition costs six months in arrears! This is fine while the subscriber base is growing but as soon as the growth rate slows down it becomes a problem. It also substantiates the Group’s contention that operators are free to finance the acquisition costs of subscribers today from the earnings from subscribers acquired using unfair cross subsidies in the past. (Which is why OFTEL was misguided in taking a 3 month snapshot for Quarter 1 of this year and disregarding the previous three year period for which it had evidence of unfair cross-subsidy.)
Extract from Vodafone Annual Report for the year ended 31st March 2000
The year saw continued rapid expansion in the UK mobile phone market, which grew by 12.4 million new customers compared with 5.8 million the previous year. There are now over 27 million mobile phone customers in the UK and market penetration is 46% compared with 26% at the beginning of the financial year. Vodafone has maintained its clear leadership in this highly competitive market-place with a record 3,216,000 net new customers, closing the year with a customer base of 8,791,000 and a market share of 32%, 5% or 1.4 million customers ahead of its nearest competitor.
Turnover in the UK increased by 39% from £2,088m to £2,901m. Operating profit, before goodwill, grew by £62m to £706m, an increase of 10%, whilst EBITDA increased by 14% to £934m. This growth in profits is after connection costs on record customer growth and continued tariff cuts.
Network business
Prepaid products have driven the growth in the UK mobile market during the year. Vodafone’s Pay As You Talk (PAYT) product has operated very successfully in this market, achieving 3,233,000 net connections in the year ended 31 March 2000, compared with 1,648,000 net connections last year. PAYT customers totalled 5,079,000 at 31 March 2000 and represented almost 58% of Vodafone’s UK customer base.
The success of PAYT is reflected in the average revenue per customer (ARPU) for the twelve months ended 31 March 2000 which, at £175 (£199 before trade discounts), was up by 10% from £159 (£178 before trade discounts) at 31 March 1999. PAYT cost to connect for the twelve months ended 31 March 2000 was held at £50 in a highly competitive market-place, compared with £43 in the twelve months to 31 March 1999.
Following a 55,000 reduction in the contract customer base in the first half of the year, revised tariffs and other changes to commercial policy resulted in a net second half increase of 38,000, giving a closing contract customer base of 3,712,000. Cost to connect rose to £94 for the 12 months ended 31 March 2000 from £88 for the comparable period, reflecting competitive pressures. ARPU was stable at £421 (£554 before trade discounts) for the twelve months ended 31 March 2000 compared to £423 (£553 before trade discounts) at 31 March 1999. This reflects tariff reductions being balanced by increased usage.
Overall average revenue per customer (both contract and PAYT) has declined from £378 last year to £305 this year due to the effect of the increase in the PAYT base. Network churn has fallen in the six months ended 31 March 2000 to 28.3% from 33.2% in the previous six months, reflecting management actions taken in the second half of the year. Overall churn in the 12 months ended 31 March 2000 rose to 29.8% from 26.0 % the previous year.,/p>
Vodafone continues to have the widest roaming capability of the UK operators, with agreements in 107 countries and across 234 networks, giving over 170 million customers access to its network. Roaming revenues, both from Vodafone customers using their phones overseas and visitors using the UK network, represented 24% of contract digital outgoing airtime and access revenues, compared with 23% last year.
Vodafone continues to invest to improve network quality. £523m was spent on capital expenditure in the year, enabling the company to sustain, and in certain areas improve, overall network quality through a period of significantly increasing demand. During the year over 1,600 base stations were installed, with 6,700 in operation at 31 March 2000.
Distribution business
The Group’s distribution companies continued to drive the majority of Vodafone’s growth, achieving net growth, excluding service provider acquisitions, of 198,000 contract customers and two thirds of the growth in PAYT. By the end of March 2000, the Group’s distribution companies accounted for 63% of the Vodafone contract customer base, up from 48% at the end of March 1999.
The share of the contract customer base connected through the Group’s distribution businesses was boosted by the acquisition of MC Mobile Services, UniqueAir, Scottish Telecommunications (Services) and 3@ Telecom during the year, for an aggregate cost of £84m.
Market leadership on PAYT has been sustained by continuing to increase availability through a wide range of retailers. Throughout the period, Vodafone has continued to work with traditional independent service providers and dealers to balance growth through these channels with that coming from new channels on PAYT.
Extract from Financial Times
FRONT PAGE - FIRST SECTION: Sales figures doubted as phones prove too mobile
Financial Times ; 06-Jul-2000
By DAN ROBERTSMore than half a million mobile phones bought in Britain are believed to have been sold abroad by people exploiting the gap between the artificially low prices of most phones in the UK and rates elsewhere.
The estimate, made by a senior executive at one of the four British mobile networks, will overshadow yesterday's claim by the industry that more than half the population now owns a phone.
Not only does the unapproved export of handsets distort true subscriber figures, but it is also costing the network operators millions of pounds in wasted subsidies.
The four companies - Vodafone, BT Cellnet, Orange and One-2-One - subsidise the retail price of almost all handsets by between Pounds 40 and Pounds 200 to accelerate subscriber growth. They hope to recoup the cost through call charges.
With normal contract plans, they have a means of monitoring customers. However, the vast majority of phones now sold in the UK use pay-as-you-go voucher schemes. These phones can easily be taken abroad without the knowledge of the network.
This is because they can be activated with a low-denomination, anonymous voucher. A purchaser could also take the phone without first activating it, but retailers would have no interest in selling it that way because they only get a subsidy from the network after the phone has been activated.
The increasing reliance on pre-pay and the fact that UK subsidies are higher than almost anywhere else in the world has made Britain a mecca for organised gangs taking lorry loads of handsets to the Middle East and other developing regions.
Vodafone, the UK's largest operator, yesterday reported having 9.36m customers. This is believed to be net of 200,000 handsets estimated to have been taken out of the country. The need to discount for missing handsets has hit network figures particularly hard in the past few months as the proportion of contract customers has declined. BT Cellnet, which claimed to add 670,000 new subscribers last quarter to a total of 8.07m, saw the number of contract customers actually fall by 151,000.
Orange, which added 1.2m to 7.17m, and One-2-One, which added 1m to 6m, now also rely heavily on pre-pay schemes to boost sales. Vodafone fared slightly better than expected with 75,000 new contract customers. Globally, it added 6.6m new subscribers.
Richard Shearer, commercial director for One-2-One, said: "While there are subsidies, and while they are higher than in other countries, this kind of fraud is in some respects a cost of doing business."
One problem is that there is no agreed standard for counting the number of true customers among the four networks. Nevertheless, the interim data posted this week show a rise of 3.44m to 30.6m subscribers at the end of June, giving the UK one of the highest penetration rates in the world.
Copyright © The Financial Times Limited