Introduction

59. A number of issues have been raised on Sky's price list for premium and other channels ('the Rate Card') and the scheme of penalties/discounts which it incorporates. To our mind the common thread is whether Sky's pricing policy constitutes the abuse of a dominant position in the wholesale market for programming supply for pay-TV (and in particular the market for premium programming) to give its retail DTH business an unfair competitive advantage over its cable competitors and is against the public interest.

60. The main questions here appear to concern:

• the level of the prices charged by Sky to the cable companies and whether these prices may be discriminatory;
  
• the general principle of the linkage between the cable wholesale price and the Sky DTH retail price and the operation of that linkage;
  
• the pay/basic ratio and other aspects of the rate card.
  

Comparative wholesale pricing

61. At the root of our unease over pricing is the fact that cable companies on average have to charge television retail prices which are in general 25 per cent higher than the Sky DTH retail prices in order to achieve a comparable gross margin - ie the margin before all non-programming costs are paid - of about 55%. An issue for the enquiry seems to be whether Sky's wholesale pricing, especially for premium programming, might be unduly discriminatory between the Sky DTH business and the cable operators.

62. The question here is whether Sky's retail business could cover its costs and make an appropriate rate of return if it had to pay the same price for its programming obtained from Sky's wholesale business as the cable companies. Accounting separation is of fundamental importance here. This question hinges on the attribution of costs and revenues between the wholesale and retail parts of Sky's business. We note that the undertakings made by Sky include an obligation not to charge the retail DTH business less than the price charged to the cable companies. This obligation does not seem however to be activated until separate accounts are produced. We understand (from Sky) that the OFT is still considering Sky's proposals for accounting separation. We have made some further comments on accounting separation in the section on suggested remedies.

The DTH-wholesale price link

63. The second pricing issue concerns the principle of the linking of wholesale price for programming supplied to the cable companies to the DTH retail price. There seem to be three aspects to this:

• the effect of the operation of this linkage given that the Sky overall package price includes third-party channels which the cable companies must pay for separately. Theoretically price increases based on adding more third party channels to the package might result in cable companies paying more for Sky programming although the Sky element of the package might be unchanged;
• the linkage between the price charged to cable operators and the overall Sky DTH package price seems to be in contradiction with one of the original objectives behind the undertakings which was to ensure that channels were charged for separately;
• taken together with the other aspects of the rate card discussed below this linkage appears to strengthen Sky's position as dominant operator and therefore market leader. Not only has it the advantage of being able to time its announcements of new services (and price increases) to best competitive effect but it takes control of the timing of price decisions out of the hands of the cable operators. Again our experience in telecoms is that the ability to the dominant operator to use price changes as a competitive weapon is important. The price linkage and other aspect also makes it more difficult to compete in offering innovative pricing and packaging option.
  

The pay/basic ratio and other aspects of the rate card

65. The pay/basic ratio is the ratio of the number of subscribers to basic packages against the number of pay subscribers (subscribers to one premium channel count as one, subscribers to two count as two and so on) - only Sky premium channels count. It is a widely-used indicator within the industry and derives originally, we understand from US industry practice.

66. Under the rate card cable operators exceeding the target for the pay/basic ratio (which is set by reference to the Sky DTH ratio) can have the price of Sky premium channels (expressed as a percentage of the Sky retail rate) reduced by up to 4 percentage points. Companies whose pay/basic ratio is below the target can have the price increased by up to 4 percentage points. This spread of 8 percentage points can make a substantial difference to cable operators' operating margins in cable television.

67. Most cable companies are still to break even on their trading costs so there is no operating margin against which to set this discount. The only company with positive earnings before interest, tax depreciation and amortisation was Videotron which achieved a margin of 10%. Clearly when companies are working very hard to become cashflow positive four percentage points either way can be of considerable importance.

68. We have no objections to the principle of Sky offering incentives for the take-up of its premium channels. This can work to the benefit of both Sky and the cable companies. However, there are two aspects of the pay/basic ratio which distinguish it from 'normal' quantity discounts. First, it can be affected solely by changes in the level of sales of third-party products - greater sales of basic packages (in which Sky channels are only one element) can affect the price charged by Sky for its programming. Secondly, there is a not insignificant penalty element. A cable operator can incur a penalty for selling a higher number of basic packages while the number of premium subscribers remains the same.

69. There appear to be strong disincentives attached to pursuing a strategy based on achieving high basic penetration but not high pay penetration (as this dilutes the pay/basic ratio). It seems reasonable to suppose that companies will factor in the effects on the price of their premium programming of signing up basic-only customers. Conceivably this could feed into higher prices for basic-only packages but equally might be a significant consideration against seeking to win additional basic-only customers by either offering enhanced value for money or low price sub-basic packages. It also seems reasonable to assume that companies would develop strategies to maximise the number of premium services taken by existing customers. Some of these (reducing the price for taking all the premium services) may be to the benefit of the consumer, but others (the withdrawal of some intermediate options) may not.

70. We have two principal objections to the pay/basic ratio. First, that it could penalise companies offering independent premium channels in competition with Sky's own offerings. There is a clear penalty if cable companies offer alternative premium programming which competes with Sky's own premium programming - not only does the cable operator have to bear the cost of the programming but also the negative effect on the pay/basic ratio where that programming substitutes for Sky's programming. Secondly that it could act as a penalty against any company which opted for a strategy focusing on low-cost basic service in order to achieve higher levels of penetration. (It is important to note that such companies might achieve the same penetration rates for sales of premium channels as companies with higher pay/basic ratios.

71. The first of these is of importance for the potential for competition in the wholesale programme supply market. The second is of immediate relevance to the ability of the cable companies to achieve higher levels of penetration. There are good grounds for believing that if cable companies are constrained into marketing strategies based on the pursuit of high pay/basic ratios this will damage their ability to achieve significant long term improvements in penetration levels.

72. A comparison with the US cable market is instructive. The average penetration achieved by the US cable operators is 56.5% with a pay/basic ratio of 72.3%. The largest operator TCI (a joint owner of TeleWest) has a 62% penetration rate with a 65.9%. TeleWest in the UK achieves a pay/basic ratio almost three times that of its US parent but its penetration is just over one third.

73. There are of course differences between the US and UK markets, but the figures from the US suggest that only a minority market segment is prepared to spend substantial amounts on television services and that the achievement of high penetration rates will be associated with 'low' pay-basic ratios - the corollary would be that strategies based on achieving high pay/basic ratios will be associated with low penetration rates.

74. It is too early to say whether the US experience will be reflected in the UK however some figures for the UK franchises are relevant:

• of the 12 franchises with penetration rates above 25% (compared to the industry average of 21%), four have pay/basic ratios of more than 150% - Enfield (Comcast operated), Thames Estuary North (TeleWest), Cotswolds (TeleWest) and Cardiff/Newport (Cable Tel);
  
• of the five franchises with penetration rates of more than 30% only one (Cardiff/Newport) has a pay/basic ratio of more than 125%.
  

75. On the question of whether there may be a linkage between high pay/basic ratio and low penetration and churn rates, it is more difficult to draw a clear conclusion. A number of factors have an impact on penetration and churn rates:

• the availability of high-quality free-to-air terrestrial channels with good reception which may mean that some consumers will not feel the need to purchase additional television services;
• the stage of network development - churn rates are inevitably higher in the early expansionary phases. Cable companies have also had to deal with the managerial implications of rapid expansion and will also freely admit that they have given highest priority to network build-out and now have to give to give greater priority to customer service systems;
• marketing strategies - initial 'taster' offers to persuade consumers to try cable services may be reflected in high levels of churn;
• the success or otherwise of strategies for managing bad debt;
• the demographic characteristics of area. In London, for example, a high proportion of people move home in any given year.
  

Finally there is the success of companies in developing anti-churn techniques such as offering sub-basic packages, using direct debits - or even free local telephone calls.

76. In view of these factors it may be that prove difficult at this moment to conclusive evidence of a link between high pay/basic ratios and high levels of churn - however Videotron's success in reducing churn rates through new pricing options, and Telecential's experience of sharply increased rates of churn do suggest that there is some linkage. Waiting until such a link can be conclusively demonstrated may mean waiting until it is too late and the damage has been done.

77. A strategy based on pursuing a high pay/basic ratio is a credible one for Sky's retail business. The pay/basic ratio for Sky's DTH business is (at 244%) significantly higher than the best achieved by any of the cable companies. 97 per cent of DTH subscribers take one premium channel and 72 per cent take all Sky's premium channels. Its churn ratio is lower. This suggests that Sky's first mover advantage has enabled it to sign up a significantly higher proportion of those consumers prepared to pay more for premium programming. It may also be that (after the high levels of churn in its early years) it has a committed and stable subscriber base. The pricing strategy it has adopted may well be the most appropriate to maximising the revenue per customer in this segment.

78. Pursuing high pay/basic ratios seems a good deal less credible as a long term strategy for the cable companies:

• they are having to win converts among the uncommitted and the pricing strategy may thus be a handicap;
• the cable industry (in contrast with Sky) is a very capital-intensive industry. The most appropriate strategy for the cable companies is to maximise the number of subscribers on the network even if incremental revenues per subscriber are small. The rate card pricing structure may operate to discourage them from doing this.
  

79. Our view is that the pay/basic ratio penalty/discount is discriminatory in that it is intended to constrain cable companies into a following a strategy based on achieving high pay/basic ratios: a strategy which reinforces the advantages enjoyed by Sky's DTH retail business while hampering the cable companies ability to increase penetration rates and minimising the scope for the cable companies to devise strategies appropriate to the capital-intensive nature of their businesses.
80. Cable networks are highly capital-intensive and the operators therefore need to find strategies to maximise the use of the network together with revenue per customer. There are differing views among cable companies about the most appropriate marketing strategies to deal with the problems of churn and low penetration. Cable companies are starting to experiment with different pricing and packaging options and the next few years will see probably see greater experimentation. It is clearly important that they should not be artificially constrained in their ability to devise such strategies.

Effects on telecommunications

81. As we noted earlier penetration rates for cable telephony are often higher than for cable television, and churn rates are often significantly lower. It is therefore difficult to gauge what the knock-on implications are for cable telephony of low penetration rates and high churn rates in cable television. Given that some 70% of subscribers take both services it seems unlikely that cable telephony would be immune from any adverse impact - although cable companies appear to be having some success in limiting the effect.

82. Pricing and packaging policies which hold back penetration rates for cable television would have a negative effect on the efficient utilisation of the network and thus on cable telephony. Revenues from cable telephony would need to fund a greater share of the fixed costs of the business as a whole.

The Basic Penetration Rate penalty/discount and volume discount

83. We recognise that in its negotiations with Sky the OFT succeeded in reducing the extent of the 'spread' of the pay/basic penalty/discount from 20 to 8 percentage points. It also increased the Basic Penetration Rate discount/penalty and Volume Discount. We believe however that further consideration needs to be given to the extent to which the Basic Penetration Rate Penalty/Discount and Volume Discount mitigate the penalty and incentive effects of the pay/basic ratio.

84. The Basic Penetration Rate is the percentage of homes passed taking cable television. The rate card offers a one percentage point discount for each percentage point by which the company's basic penetration exceeds its performance in the previous year up to a maximum of 3 percentage points. There is also a one percentage point penalty for every percentage point by which it falls short of this baseline up to a maximum of 3 percentage points.

85. We note that the maximum basic penetration discount is 3 percentage points whereas the maximum pay/basic penalty is 4 percentage points - so a company with a low pay/basic ratio might still incur a penalty (of one percentage point) even if it achieved the maximum penetration discount.

86. There seem to be good reasons why the PBR may not have an incentive effect to mitigate the penalty effect of the pay/basic ratio. The main reason is that, given that the baseline in each year would be the company's performance in the previous year, in order to keep the discount the company must exceed its previous performance. Companies are wrestling with the problem of how to increase penetration above the industry average BPR about 21 per cent. In this context the prospect of sustaining year-on-year gains is uncertain and the prospects of sustaining year-on-year gains of 3 percentage points still more uncertain; indeed, the risk for companies is that exceptional performance in one year might bring with it the risk of falling short against the increased benchmark.

87. Under the Volume Discount aspect of the Rate Card a 3 percentage point discount is available for companies with more than 250,000 subscribers; 2 percentage point for 150,000 to 249,999; and 1 percentage point for 100,000 to 149,999. Again however there are grounds for doubting the incentive/penalty effect: only one operator would qualify for the full 3 percentage points, and it seems unlikely that a company would be able to achieve a gain of more than one percentage point through organic growth - at least in the short term.

88. Irrespective of the question of how many operators would qualify levels for what levels of discount and when, it is open to doubt to what extent this discount does actually counteract the incentive/penalty effect of the pay/basic ratio - even if a company qualified for the full 3 percentage points it would still have strong financial reasons for wanting to maximise its pay/basic ratio.

Conclusion: the pay/basic-ratio penalty/discount

89. We are dependent on the published information on the rate card so it is difficult to make a judgement on the underlying rationale for its structure and the choice of the levels of discount. Concerns must however remain about the effects of the pay/basic ratio penalty/discount. If the analysis set out above is correct then these effects may not be mitigated to any substantial extent by the penetration and volume discounts.

90. We recognise that the severity of the effects of the pay/basic penalty/discount were reduced. However, in our view they still remain significant.

Channel carriage discount

91. The rate card offers a discount of 3 percentage points if all Sky channels are carried. This is reduced by 0.2 percentage points for each channel which is not carried.

92. We note that in its negotiations with Sky the OFT succeeded in dealing with the negative pricing implications of the original proposals. That said the principle of an 'all-channels' discount remains of concern and this could emerge to have greater significance as Sky launches more channels. The capacity of cable systems is only just starting to emerge as an issue - cable companies are now only just starting to be in the position where they are able to choose between programme suppliers rather than take what they could get. It is fully possible - given the large amounts of money that cable operators spend on Sky programming that this discount will make a difference in these decisions. It is not necessarily the case that the introduction of digital technology will reduce the significance of capacity constraints - and indeed the potentially anti-competitive effect of such discounting might be magnified if it were to be extended to Near Video On Demand which involves a large number of channels.

Conclusion: the rate card

93. The preceding sections have raised a series of questions and concerns about the structure of the rate card and the specific numbers attached to the various incentives/penalties. Our view is that the cumulative effect of these concerns is such as to highlight the need for a fundamental reconsideration of the card. We have made suggestions on alternative options for consideration in our section on suggested remedies.

94. We would also note that while the rate card includes elements which might be deemed to be compensating the cable companies for effects of the pay/basic ratio penalty/discount such elements do not however compensate third parties (such as competing channel providers) who may experience loss as a result of the effects of the discount structure on cable companies pricing and packaging decisions.

Possible remedies

95. Ensuring non-discriminatory pricing: In our view, given Sky's position in both the market for the wholesale supply of programming and the retail-distribution market, non-discriminatory pricing based on accounting separation is a key safeguard against abuse and should be a central part any undertakings. Oftel has considerable experience in this area and would be happy to assist. Our experience in telecommunications suggests that for it to be effective accounting separation should involve:

• the publication of the accounts at regular intervals (every 6 months). We believe that publication of the accounts ensures that the accounting separation arrangements are subject to informed scrutiny by the industry a process which yields essential information to the regulator;
• independent audit to regulatory standards;
• clear ground rules on the attribution of costs;
• clear transfer charges;
• provision for modification of the arrangements in the light of experience.
  

The attribution of costs and revenues is a critical issue. The issues here appear to us to include treatment of:

• transmission costs;
• sales and marketing costs;
• advertising revenue.
  

96. We would note on this last point that given the amount of these revenue their attribution could make a significant difference to the outcome of this process. Since Sky's wholesale business precludes the cable companies from competing with Sky's retail business for advertising revenue it would be difficult to regard this as a characteristic inherent to Sky's retail business and not to the cable companies cable television business. There appears to be a good argument for taking this into account in the attribution of costs and revenues. If this revenue is not attributed to the wholesale business then there would be a case for regarding it as an exclusive right granted to Sky's retail business and for taking it into account in assessing whether the prices charged to Sky's retail business and the cable companies were genuinely non-discriminatory.

97. We believe that the obligation not to discriminate in favour of the retail business should not be dependent on the production of separate accounts.

98. Linkage between wholesale price and DTH price: Our view is that non-discriminatory wholesale prices per channel (based on a proper distribution of costs) provides the right basis to go forward. This would also provide a baseline against which issues of predatory pricing might be judged should they ever arise.

99. Discounts: our view is that at the very least the incentive arrangements should not involve a penalty for promoting the sale of basic-only services. We would suggest that consideration be given to discounts relating to the levels of premium subscribers - expressed either in absolute terms or as a percentage of homes marketed (as distinct from basic sales). We believe that a discount per channel would be most in line with the objective of preventing channel bundling - we would not support the continuation of a channel carriage discount.

Go to next section of this document