Final Report for Department for Business, Innovation and Skills and Department for Culture, Media and Sport
Download 2.13 Mb.
|
- Navigate this page:
- Producer surplus assumptions and forecasts
Consumer surplus calculations Consumer surplus is calculated separately for mobile voice, handset data and mobile broadband, using the formula: 
with the relevant ASPU, choke price and number of subscribers for each service. We have calculated annual values for 2011 so that they can be compared with the findings of the 2006 Europe Economics study. We have also compared our figures against the results of the 2006 study, updated for the change in RPI between 2006 and 2011. Finally, we provide calculations of the net present value of the cumulative consumer surplus over a ten-year period from 2012 to 2021. We chose a ten-year period for the longer-term view because it approximates to the length of previous mobile technology cycles.124 The NPV is calculated by assuming a social discount rate of 3.5% per annum. Producer surplus assumptions and forecasts Revenue Figure B .13 shows the evolution and breakdown of total mobile service revenue; this is calculated by considering the ASPU and number of subscribers, as presented earlier. In addition, we then make an allowance for other revenue, such as messaging and interconnect. We assume that total revenue will stay relatively constant in nominal terms, as operators begin to offer new value-added services125 in order to offset diminishing revenue from basic voice, messaging and packet data services. Figure B.13: Total mobile service revenue – historical values and forecasts [Source: Analysys Mason, 2012] 
Number of sites In order to estimate network costs, it is necessary to forecast the number of mobile sites over the next ten years. We have reasonably current information for this, sourced from the Ofcom ‘Sitefinder’ database. We understand that the number of Everything Everywhere sites is not up to date on this database; since Everything Everywhere is in the process of consolidating existing Orange and T-Mobile sites we have assumed that the current number of sites is 95% of the figure quoted in Sitefinder. We do not have historical information for the total number of sites in the UK, although we are aware that the number has been decreasing (again due to the consolidation of Orange and T-Mobile sites by Everything Everywhere). We have used data from the Ofcom mobile long-run incremental cost (LRIC) model126 to estimate the proportions of macro-cell sites and small-cell sites and the technology mix at each type of site. The resulting assumptions are shown in the following figures.
Coverage sites We have made an assessment of the number of 800MHz 4G coverage sites likely to be required, using data from Ofcom’s 2012 consultation on the 800MHz and 2600MHz spectrum award.127 We have also used this information to estimate the number of 900MHz sites required by O2 and Vodafone, and 1800MHz sites required by Everything Everywhere, to maintain current levels of 2G coverage. We believe it is likely that a large proportion of the base stations required for the 4G coverage network can be deployed at existing sites. So far, we have assumed that all of the 4G coverage networks are rolled out using 800MHz spectrum, although we recognise that in practice one operator may be forced to roll out a 4G coverage network at 1800MHz. In addition, we assume that the level of inter-network tower sharing will increase significantly over the next few years, since Three and Everything Everywhere have an agreement through the tower company MBNL while Vodafone and O2 have an agreement through the tower company Cornerstone. We assume that because of these tower sharing agreements, virtually all 4G coverage towers will be shared by two operators, and significant consolidation of 2G and 3G sites will take place in the next few years. Capacity sites The number of capacity sites is determined by considering the following three factors: number of subscribers traffic per subscriber (extracted from the Ofcom LRIC model) capacity sites required per busy-hour Mbit/s (extracted from the Ofcom LRIC model). We have assumed that capacity per site will increase over time, due to technological advances which result in more-efficient use of existing spectrum (the migration to LTE-Advanced, for example). Traffic per subscriber is assumed to grow at the rates predicted by Analysys Mason Research’s Wireless Traffic Forecast, shown in Figure B .16 below. We note that other third-party forecasts, such as Cisco’s Visual Networking Index,128 assume faster growth in traffic per subscriber (Cisco’s forecast, converted into traffic per subscriber using Analysys Mason’s subscriber forecast, is also shown in Figure B .16 for comparison purposes). The report therefore includes a sensitivity based on the Cisco forecast, although this appears to show that this level of traffic growth is inconsistent with our longer-term revenue assumptions, since the producer surplus becomes negative in 2021. Figure B.16: Data traffic per device – historical values and forecasts129 [Source: Analysys Mason, Cisco, 2012] 
As with coverage sites, capacity sites are calculated separately for each technology, considering the number of incremental sites required to fulfil the demand of 2G, 3G or 4G subscribers. We have taken into account both technology site sharing and also inter-network site sharing by considering the effect of Cornerstone and MBNL. Costs In order to estimate the producer surplus for mobile broadband operators we have modelled the capital expenditure (capex), the operating expenditure (opex), and the costs of goods sold (CoGS). The cost inputs are based on annual reports from UK operators supplemented, where necessary, with data from Analysys Mason internal databases. Capex The majority of capex in a mobile network is attributed to rolling out new base stations (coverage and capacity sites, as described previously), in addition to investment in core network equipment, and replacement capex (replacing equipment at the end of its lifecycle). To reflect this in the model, capex is primarily driven by the number and type of sites, and the number of radios per site. For 4G this involves rolling out a coverage network between 2013 and 2015, followed by incremental capacity sites. For 2G and 3G, sites are rolled out or decommissioned based on capacity requirements, although a minimum 2G coverage network is assumed to remain in place during the forecast period. The 4G network can be rolled out by either deploying new 4G base stations, or upgrading existing base stations to 4G, which as mentioned above, is likely to be the most common case. The capex model thus considers four main types of cost: set-up costs for the main sites upgrade costs for the main sites capex for the micro sites replacement costs. Capex has been calculated separately for 2G, 3G and 4G networks. Figure B .17 below shows the capex in each year for all operators and across 2G, 3G and 4G networks. Figure B.17: Capex over the forecast period [Source: Analysys Mason, 2012] 
Opex The opex model considers eight types of cost: staff costs site running costs (including backhaul, electricity cost) site maintenance costs (including equipment) microsite running and maintenance costs site rental costs marketing costs general and administrative costs bad debt. Opex has been calculated separately for 2G, 3G and 4G networks. Figure B .18 below shows the opex in each year for all operators and across 2G, 3G and 4G networks. Figure B.18: Opex over the forecast period [Source: Analysys Mason, 2012] 
CoGS CoGS consists primarily of subscriber acquisition and retention costs (dealer commissions, equipment subsidies and distribution expenses). Note that we have not included interconnection revenue or costs, as for the most part these are passed between the operators within a single country and, as such, do not have an impact on the direct economic welfare for the country as a whole. We assume that CoGS accounts for 30% of total revenue throughout the modelling period, and this applies consistently across both mobile broadband and handsets. Equipment subsidies, and in particular dealer commissions, are kept confidential by the operators and are difficult to benchmark. In addition, different operators tend to group costs into different categories. We have therefore taken a bottom-up approach of modelling network and non-network opex in detail, and have then estimated the CoGS by considering operators’ total opex and cost of sales.130 Directory: government -> uploads -> system -> uploads -> attachment data -> file file -> Voluntary and mandatory surrenders 2013 file -> Tensions in Sino-Japanese Relations in 2012 and 2013: Plus ça change? file -> Review of Gaming Machine Stake and Prize Limits file -> E-bulk rb system Test Plan commercial contents file -> Consular section file -> Documents that can be used as proof of address and identity file -> Positive for Youth Discussion Paper June 2011 Capital infrastructure to support services for young people file -> Open Source Software Options for Government Version 0, April 2012 Aim file -> Aviation House 125 Kingsway Download 2.13 Mb. Share with your friends:
|
