7A new broadband utility?
172.Although functional separation was a key outcome of Ofcom’s 2005 telecommunications review, Ofcom is now concerned that this model has failed sufficiently to counteract the embedded conflict of interest between BT and Openreach, and it has concluded that further reform is required. Ofcom believes that Openreach must behave like, and be seen to behave as if it were, an independent company. Ofcom has indicated that full structural separation, in which Openreach would be spun off as a separately quoted company, has not been ruled out. In evidence to us, Professor Dieter Helm, a prominent economist specialising in utilities, infrastructure and regulation, made a robust challenge to the status quo in regard to the UK’s communications market. Professor Helm argued that the only way Openreach can “behave like” an independent company is for it actually to be an independent company.217 He saw the DCR as offering a key turning point for change, but feared that Ofcom is baulking at making a difficult decision and failing to follow through on the logic of its own analysis.
173.Professor Helm’s view is that Openreach’s status within BT creates inherent conflicts of interest in favouring BT over other users of its network, and that the regulator will find it ever more difficult to regulate Openreach effectively. Were Openreach to be an independent company, he believes it should be regulated as a utility; it would then benefit from a significantly lower cost of capital, leading to more investment in infrastructure and potentially lower broadband costs for consumers. Such a company, in his opinion, would be a natural candidate for “managing” a universal service obligation, even if not necessarily providing all the infrastructure or investment itself.
174.According to Professor Helm, under a regulated asset base (RAB) model Openreach should become the overall broadband system operator, be licensed to develop and run national broadband infrastructure as a form of utility. 218 In return, it would be provided with a guarantee that its licensed functions would be financed. To do this it would levy a use-of-system charge across its entire customer base. Its resulting cost of capital would be low, given the financial protections in place, and it could with confidence leverage private funding to complete and upgrade the broadband network.
175.SSE, an energy company, also identified some advantages of a utility-style approach for telecoms.219 SSE explained that a utility approach for broadband could allow infrastructure to be financed more efficiently by a greater range of potential investors, whose confidence to invest would be supported by greater understanding of how new discrete networks could be absorbed into a logical utility whole. Prospective providers of superfast broadband infrastructure should be able to “plug in” their network to a national access network that provides a platform through which any retail service provider could supply their services to end customers on the new network on an open-access basis.
176.Nonetheless, there are marked differences between the communications market and those of electricity, water and gas. In broadband and telecoms, in direct contrast to many utilities, the original incumbent operator BT faces infrastructure competition at several levels of the network; and service competition, as multiple communications providers compete to supply telecoms services to consumers by reselling services or via use of Openreach’s access network.220 In providing local access infrastructure BT competes with Virgin Media for around 45% of the country’s households today and Virgin plans to extend its network to reach around 60% of households. There are many other smaller access network providers, for example Gigaclear and Hyperoptic.221 Moreover, Sharon White told us that she saw the possibility in the longer term for 40% of households to benefit from competition in local access infrastructure where Openreach’s infrastructure would be subject to competition from at least two challengers (e.g. Virgin Media and another provider).222 A further difference is that the communications market is characterised by a quicker pace of technological change, where the asset lives associated with infrastructure are much shorter than in other utilities. There is thus a greater risk of costs being stranded and not being recoverable.
177.In fact Openreach operates under a near-RAB regime at present. Ofcom told us that a more fully RAB model could be used in certain circumstances to drive investments in a particular asset but that it saw problems with this approach. 223 It appears to remain Ofcom’s view that it is better to use competition to drive investment in network upgrades and a fully RAB approach would conflict with competition in the access network.224 It could also potentially allow exploitation of market power by a dominant network operator to driver higher prices, making customers who do not value a service upgrade still pay for it.
Incentive to invest in national infrastructure
178.Like any company that wishes to raise outside finance, BT faces a cost of capital from the market, which is a function of its costs of debt and equity. BT confirmed to us that Openreach does not have its own capital structure or debt. Financing is raised at BT Group level rather than at the Openreach level. In assessing all projects in Openreach and for other parts of the BT Group, BT uses a group cost of capital, which is currently 10.4% (nominal pre-tax rate).225 BT Group’s average weighted cost of capital has most recently been estimated by Ofcom to be 9.9%. Ofcom then disaggregates this estimate to reflect the activities and risk profiles covered by Openreach and the rest of BT.226 It has recently calculated Openreach’s cost of capital to be 8.8%, and that of the rest of BT to be 12.4%.
179.There was some debate about the effect that structural separation would have on Openreach’s costs of capital. Sean Williams, BT’s managing director of strategy, portfolio policy, argued that Professor Helm’s assertion that BT’s cost of capital would be cheaper for Openreach outside BT was completely wrong.227 According to him, a separate Openreach would be seen by investors as more of a risk than it is today, because the new, smaller company would not be able to spread risks across the wider Group. Professor Helm argued that the opposite would be likely.228 He also stated that BT would not, in practice, invest less at Openreach, if it were forced to hive off the operation, comparing the situation to the past break-up of British Gas—where he maintained that there had been no such investment effects. Apart from there being clear differences between the two industries, BT has strongly warned about the disruption any break-up might cause. This is an issue which the Minister also took seriously.
180.Sharon White suggested that whether a structurally separated Openreach would invest more than it does at present was still an open question, but that Ofcom’s ultimate objective was to see more investment in the network.229 As the regulator, Ofcom aims to incentivise BT to make efficiencies so that increased investment goes into the repair and resilience of their network. The extent to which this has been happening is questionable.
|
Weighted average cost of capital comparison |
||||
|
Openreach |
Openreach |
Ofgem3 |
Ofwat3 |
|
|
Pre-tax nominal1 |
Real vanilla2 |
Real vanilla |
Real vanilla |
|
|
Cost of equity |
10.2% |
4.8% |
6.0% |
5.65% |
|
Cost of debt |
5.4% |
2.1% |
2.6% |
2.6% |
|
Gearing |
30% |
30% |
65% |
62.5% |
|
WACC |
8.8% |
4.0% |
3.8% |
3.7% |
Source: Ofcom
Note: 2016 Openreach WACC vs 2014 Ofwat/Ofgem WACCs
1. The pre-tax nominal rate of return is the rate of return that an investor receives before taking off any taxes paid by the investor and without adjustment for inflation. The real rate of return is the rate after adjustment for inflation. For example, if the nominal rate of return is 5% and inflation is 2%, then the real rate of return is 3%.
2. The vanilla WACC is estimated by reference to the post-tax cost of equity and the pre-tax cost of debt, while the pre-tax WACC is estimated by reference to the pre-tax cost of equity and debt
3. These rates are applied by Ofgem and Ofwat to those parts of the energy and water sectors which they regulate.
181.From the information we have seen on other utilities’ capital costs, it would appear that they do attract costs of capital well below almost any other asset outside of government bonds.230 Similarly, Openreach’s estimated weighted average cost of capital231 is only a little higher than that of the utilities above. Utilities are particularly attractive investments to institutional investors such as international pension, life assurance and wealth funds, who are looking for the certainty of return and low risk that they represent. From Ofcom’s competition-based perspective, it did not consider that the primary driver of the cost of capital was structure, but rather the financial risk of BT’s activities, i.e. the extent to which BT’s return on capital was guaranteed.
182.BT provided us with a survey it commissioned which suggested, among other things, that more than half of the institutional investors surveyed thought that cost of capital (58%) and the level of risk of investing in Openreach (56%) would increase if Openreach were structurally separated from BT Group.232 However, the survey predominantly canvassed existing BT investors, 82% of those surveyed, and provided no indication whether these represented a cross-section of investors overall.233 It is not clear whether BT currently attracts the same proportion of long-term institutional investors as would be reflected in the capital and equity markets as a whole. Finally, as Ofcom has indicated, BT is earning above-market returns and appears to benefit financially from the conflict of interest between its wholesale and retail businesses. If these advantages were not maintained following a split, that would create a financial incentive for current investors to prefer the current model to structural separation.
183.Given that 90% to 95% of Openreach’s activities are regulated by Ofcom and subject to price caps, it is already fairly close to a regulated asset base model. Yet, as the Committee’s advisory panel has highlighted, the fact that Openreach has to pass BT’s investment “hurdle” rate, which is set at the BT Group weighted average cost of capital (10.4% by BT’s estimate), while Ofcom allows it a rate of return at 8.8% suggests that Openreach is presently under-investing in its infrastructure and business. This also suggests that the utility-like part of the Group, Openreach (sitting inside BT), is in effect subsidising riskier projects in the rest of the Group such as sports rights and IP TV channels. By requiring a higher rate of return (10.4%) than the estimated cost of capital for Openreach (8.8%) BT will, inevitably, reduce investment in Openreach to below optimal levels in this part of the business. Profitable projects whose expected return falls between these rates will not be pursued, to the likely detriment of shareholder value. The profile of BT’s potential investments is not clear, but the amounts of profitable infrastructure investment forgone as a result of the 1.6% gap between hurdle rate and cost of capital could potentially be hundreds of millions of pounds a year. Shareholders, as well as BT’s customers, therefore, should welcome higher levels of investment in the local access network. Also, in the long run, BT’s investment approach will push up the Group’s cost of capital, as riskier areas will over-expand and safer areas invest too little.
184.In relation to its use of a group hurdle rate to guide investment decisions, BT told us that it was not true that the Group utilises some of the returns from Openreach in an incorrect manner. BT explained that all the costs of competitive services such as sports rights were both accounted for in the retail business and funded by retail charges. However, even where Openreach’s products are under a charge control regime, it is not clear from BT’s accounts what contributions each division’s free cash flow makes to wider Group spending and to the company’s dividends to shareholders. It could be the case that Openreach is shouldering a disproportionate amount of these costs. In fact, Ofcom has confirmed that following a review of the way BT allocates its costs to regulated services, it estimated that BT had overstated costs in its 2014/15 regulatory financial statement (and therefore underestimated its returns) by £225m.234 This shows the ability of BT to exploit the Group’s vertical integration with Openreach and potentially to manipulate Ofcom’s charge controls and market reviews to its advantage.
185.Communications providers who rely on the Openreach’s infrastructure argue that BT has not invested enough of the money it makes from the Openreach network back into the network. Although no-one has calculated the quantum of the underinvestment, this claim does correspond to the way we discovered that BT allocates capital between its various businesses on the basis of a single group-wide hurdle rate. We consider this is bound to lead to sub-optimal investment in the relatively low-risk Openreach.
186.There appears to be compelling evidence that BT Group is exploiting the position of vertical integration to make strategic decisions that favour the Group’s priorities and interests, at the expense of its access infrastructure business. BT does not lack access to capital. Its current structure allows it to use Openreach’s utility-type assets to cross-subsidise riskier activities elsewhere in the Group, while significantly under-investing in Openreach.
187.It came as a surprise to us that BT employs an investment hurdle rate significantly above Openreach’s actual cost of capital, as estimated and allowed for by Ofcom. At the same time, BT’s use of an investment hurdle rate which is 1.6% above Openreach’s cost of capital means that a potentially very significant amount of annual investment in broadband access and services, investment that would add to shareholder value, is not made. While we understand the desire for BT and other providers to balance infrastructure investment with their own commercial interests, this forgone investment in maintaining, upgrading and supporting Openreach’s infrastructure is damaging both to public welfare, to shareholders and to consumers. We believe there is a pressing need to liberate more of Openreach’s revenue for investment in broadband and the evolution of its telecoms infrastructure. As a result there is a need to consider closely BT’s governance and capital structures as well the adequacy of its oversight and regulatory arrangements.
188.On the evidence presented, it seems very likely that Openreach would invest more in upgrading its infrastructure if it were a separate company, since it would not be competing with other Group businesses and would be freed from the Group hurdle rate on investment. By adopting its current approach, BT is likely to be sacrificing shareholder value and public benefits that would flow from these investments. This is likely to mean that substantial amounts of money—potentially totalling hundreds of millions of pounds a year—are not being invested in developing and upgrading Openreach infrastructure which is critical to the UK economy and most people’s lives. We therefore recommend that Ofcom undertakes an assessment to ascertain the financial effect of BT’s failure to invest in Openreach at its true cost of capital.
218 The regulatory asset base (RAB) usually refers to the measure of the net value of a company’s regulated assets used in price regulation.
220 There are infrastructure-based competitors at several stages of the network: e.g. Sky & TalkTalk have invested in infrastructure as far as most of BT’s exchanges.
222 [Q946]
224 In a RAB model the regulator guarantees the regulated firm a return on its assets, which would be extremely difficult to do in a setting where those assets may be rendered obsolete by a competitor taking away the business when a customer switches from one network to another’s, for example from Openreach to Virgin or Gigaclear, who use their own infrastructure and not Openreach’s.
229 Q911
231 Weighted average cost of capital (WACC) is a calculation of a firm’s cost of capital in which each category of capital is proportionately weighted. All sources of capital, including stock, bonds and any other long-term debt, are included in a WACC calculation.
232 The survey was carried out by Brunswick Insight for BT and completed in April 2016.
233 71% of respondents were based in the UK and were made up of a mix of telecoms buy-side analysts, generalist fund managers and telecoms fund managers.
© Parliamentary copyright 2015
18 July 2016