Spectrum Workshop - Issues and challenges in spectrum allocations and spectrum valuation

50 %
50 %
Information about Spectrum Workshop - Issues and challenges in spectrum allocations and...

Published on March 10, 2014

Author: StefanZehle

Source: slideshare.net


Stefan Zehle, CEO Coleago Consulting discusses issues and challenges in spectrum allocations and spectrum valuation. Find out more on www.Coleago.co.uk

Spectrum workshop Issues and challenges in spectrum allocations and spectrum valuation Stefan Zehle, CEO Coleago Consulting stefan.zehle@coleago.com +44 7974 356 258

2 Introductions Market, spectrum and technology overview Principles of spectrum valuation Spectrum valuation in practice Spectrum renewal Summary Appendix 1 - Steps in Network Consultation Appendix 2 - The Modelling Process Agenda © Copyright Coleago 2012Spectrum Workshop

Our skills and experience Introductions 3© Copyright Coleago 2012Spectrum Workshop

Coleago offers advisory services focused on the TMT sector 4 Strategy & Business Planning Marketing & Customer Management Regulation & Interconnect Business Transformation & Cost Reduction Due Diligence Improving Network Performance Spectrum and Licences Digital Content & Media Training © Copyright Coleago 2012Spectrum Workshop

A refreshingly different approach to consulting 5 Junior Consultant Senior Consultant Manager Senior Managers Partner Analyst Traditional Consulting Firm Model © Copyright Coleago 2012Spectrum Workshop

Spectrum related work 6 Country Year France 1994/5 Hong Kong 1995 Singapore 1995 Poland 1995 Austria 1995 Belgium 1995 Finland 1996 Denmark 1996 Italy 1996 Taiwan 1996 Brazil 1996/7 Mexico 1996/7 Argentina 1997 Country Year Czech Republic 1999 Netherlands 2000 United Kingdom 2000 Belgium 2000 Egypt 2006 US (AWS) 2006 Canada (AWS) 2008 India (3G & BWA) 2009/10 UK (2.6GHz) 2009 Poland (2.6GHz) 2009 Ukraine (3G) 2010 Thailand (3G) 2010 Ireland (Renewal) 2010 Country Year Canada 2001 Ireland 2002 Sudan 2003 Algeria 2004 Iran 2004 Maldives 2004 Oman 2004 Saudi Arabia 2004 USA 2006 Belgium 1997 Kuwait 1997 Ireland 1997 Norway 1998 © Copyright Coleago 2012Spectrum Workshop

2011 was an interesting year and 2012 will be similar! 7 Country Year Australia renewal 2011/12 Bangladesh 3G 2011 New Zealand 2011 Switzerland 2011/12 United Kingdom 2011/12 Syria 2011 Belgium 2011 Pakistan 2012 © Copyright Coleago 2012Spectrum Workshop

Our spectrum expertise is much broader than simply spectrum modelling and valuation Spectrum Consultations Auction Strategy & Auction Readiness Administered Incentive Pricing Live Auction Support © Copyright Coleago 2012Spectrum Workshop 8

The role of spectrum in determining operator performance Market development, spectrum and technology overview 9© Copyright Coleago 2012Spectrum Workshop

Significant traffic growth driven by mass market adoption of smartphones and large-screen mobile broadband 10 Step Change in Mobile Broadband Data speeds capable of delivering a good user experience Data optimised and desirable devices for consumer and business applications Flat fees; pricing consistent with the application Compelling applications and content Time Data Traffic Minimum Acceptable User Experience 0 1 10 100 LTE 5 50 HSPD A+ CSD HSCS D GPRS HSPD A Mbit/s Time EDGE - 10 20 30 40 50 60 70 80 90 3 O2 Orange Vodafone T-Mobile GB£/GByte Jun-06 May-08 Mar-09 © Copyright Coleago 2012Spectrum Workshop

Mobile data traffic is not just about laptops and smartphones… 11 Cisco forecasts imply an average data usage CAGR of ~70% p.a. © Copyright Coleago 2012Spectrum Workshop

Mobile data traffic is rapid and consumer create most of the traffic, example Sweden © Copyright Coleago 2012 0 50 100 150 200 250 300 2H 061H 072H 071H 082H 081H 092H09 Mbytes/Customer/Month Mobile Data Usage in Sweden Consumer Business Blended 0 5,000 10,000 15,000 20,000 25,000 30,000 1H 07 1H 08 1H 09 Terrabytes AnnualMobile Data Traffic in Sweden Private Business Total 12Spectrum Workshop

Singapore exhibits similar trends as see in Europe 131 244 550 802 1,007 1,615 2,197 2,626 0 500 1,000 1,500 2,000 2,500 3,000 1Q08 2Q08 3Q08 4Q08 1Q09 2Q09 3Q09 4Q09 TbytesperQuarter StarHub Singapore PacketData Traffic 102 164 327 439 514 779 1,005 1,141 0 200 400 600 800 1,000 1,200 1Q08 2Q08 3Q08 4Q08 1Q09 2Q09 3Q09 4Q09 Mbytes/3GUser/Month StarHub Singapore PacketData Traffic per 3G Customer © Copyright Coleago 2012 13Spectrum Workshop

Technology developments: LTE is now a commercial reality and WiMAX is potentially being marginalised 14 FDD TDD HSPA LTE WiMAX LTE 200+ operators support 3.6Mbps+ circa 40 operators support 14.4Mbps The first commercial LTE service launched December 2009 in Sweden and Norway by TeliaSonera Test networks launched at 800MHz and 2.6GHz in Germany (Sep 2010) LTE 1800 trials in Finland and Australia Russian telecoms operator Yoto (2nd largest worldwide WiMax player) intends to switch from WiMax to LTE for its future roll-out. Clearwire, the largest WiMax player, and Sprint, have also made statements that neither would rule out LTE technology. Nortel’s Chinese R&D unit completed a streaming video session over LTE TDD. Likely that LTE rather than WiMax will be rolled out in India. MWC Barcelona: TDD LTE equipment 1-2 years from mass production. © Copyright Coleago 2012Spectrum Workshop

0% 5% 10% 15% 20% 25% 30% - 10 20 30 40 50 60 4Q05 1Q06 2Q06 3Q06 4Q06 1Q07 2Q07 3Q07 4Q07 1Q08 2Q08 3Q08 4Q08 1Q09 2Q09 3Q09 4Q09 ARPU tarHub Singapore - Voice & Data ARPU Voice Data Data % Decoupling of traffic and revenues: growth in data usage far outstrips data ARPU growth 15 0% 5% 10% 15% 20% 25% 30% 0 5 10 15 20 25 30 35 40 1Q06 2Q06 3Q06 4Q06 1Q07 2Q07 3Q07 4Q07 1Q08 2Q08 3Q08 4Q08 1Q09 2Q09 3Q09 4Q09 Data%ofARPU MonthlyARPU€ Orange France Data Voice and Data ARPU Voice Data Data ARPU % 0% 5% 10% 15% 20% 25% 30% 35% 0 10 20 30 40 50 60 Q107 Q207 Q307 Q407 Q108 Q208 Q308 Q408 Q109 Q209 Q309 Q409 Data%ofARPU MonthlyARPUUS$ T&T USA - Voice and Data ARPU Data Voice Data % of ARPU MonthlyARPUS$ Orange France voice and data ARPU 0 10 20 30 40 50 60 Q107 Q207 Q307 Q407 Q108 Q208 Q308 Q408 Q109 Q209 Q309 Q409 MonthlyARPUUS$ AT&T USA - Voice and Data ARPU Data Voice Data % of ARPU  Data revenue growth (partly) offsets declining voice ARPU  Relative to data traffic however, data revenue growth remains modest  Significant investments relative to revenue StarHub Singapore voice and data ARPU MonthlyARPU€ © Copyright Coleago 2012Spectrum Workshop

More spectrum becoming available for mobile communications 16 Mobile DigitalTV Digital Dividend Mobile GSM 3G 3G 3G LTE Mobile Expansion 450MHz 470MHz 698MHz 806MHz 850MHz 900MHz 1800MHz 1900MHz 2100MHz 2300MHz 2500MHz 3400MHz 3600MHz Better for Coverage Geographic Coverage Indoor Coverage Better for Capacity © Copyright Coleago 2012Spectrum Workshop

The cost of geographic coverage is lower with lower frequency bands 17 100% 126% 323% 455% 675% 1230% 0% 200% 400% 600% 800% 1000% 1200% 1400% 700MHz 850MHz 2100MHz 2500MHz 3500MHz 5800MHz Relative Capex to Build Coverage Coverage of rural areas with digital dividend 700/800 MHz spectrum is about 30% of the cost compared with 2100 MHz coverage. Source: SFR, France © Copyright Coleago 2012Spectrum Workshop

 The superior propagation characteristics of low band spectrum yield a higher probability of (deep) in-building service at a given performance  This may be translated in terms of a % of demand that is only addressable with low band spectrum (or with potentially expensive in-building solutions) Indoor propagation characteristics 18 Probability of service With spectrum below 1GHz With spectrum above 1GHz % DemandDeep indoor © Copyright Coleago 2012Spectrum Workshop

The surge in data traffic results in a incremental capex and a squeeze on margins  The capex-to-sales ratio is increasing again, reversing the declining trend – Threat of negative returns on incremental customers  In response to high traffic volumes, operators have started to abandon unlimited data plans in 2010  Increased prevalence of service restrictions (e.g. P2P, continuous video) – However, future net neutrality regulations could preclude this 19 Congestion squeezes margins Time Revenues per incremental customer decline Marginal costs per customer increase at capacity limits Squeeze on margins due to costs of congestion-relief © Copyright Coleago 2012Spectrum Workshop

Shifting emphasis: Is capacity to data what geographic coverage was to Voice?  As networks ‘fill up’: – Capacity becomes a scarce resource – Mobile communications shifts from a “buyer’s market” to a “seller’s market”  Key premises: – It may not be feasible to build oneself out of congestion – Traffic will tend to migrate from more congested networks to less congested networks  If so, market shares will tend to converge towards shares of industry capacity – Direct impact: market shares of data traffic and revenues – Indirect impact: market shares of voice, through linked / correlated accounts  Maintaining a fair share of industry capacity is of strategic importance 20© Copyright Coleago 2012Spectrum Workshop

There may be no alternative to securing spectrum  But securing fair-share of capacity is a strategic imperative for individual operators – To ensure a level playing field and to protect (or grow) market share  Once other means have been exhausted, spectrum acquisition may be the only remaining option 21 Shares of Spectrum and capacity Share of Industry Capacity MNO average MNO Amount of Spectrum acquired by an MNO Minimum required for capacity parity Demand vs. Industry capacity Time Capacity Market Congestion? © Copyright Coleago 2012Spectrum Workshop

Spectrum is moving centre stage  GSM Spectrum liberalisation / re-farming – Introduction of technology neutrality, enabling introduction of HSPA and LTE in former GSM bands  New Mobile Spectrum releases – 2.6GHz and the digital dividend spectrum in 700 / 800 MHz  Licence Expiry – In many countries the 900 and 1800 MHz GSM licences are near to expiry, and some 2.1 GHz licence are only 4 years away from expiry – Renewal may involve negotiations around Administered Incentive Pricing New mobile spectrum releases and spectrum renewal negotiations invariably call for a spectrum valuation exercise, in which all existing and expected future spectrum needs to be considered jointly 22© Copyright Coleago 2012Spectrum Workshop

The basic principles of valuing spectrum Spectrum valuation 23© Copyright Coleago 2012Spectrum Workshop

Spectrum valuation is central to spectrum award and renewal processes Shareholder Value What and how much spectrum do we want? How do we obtain the spectrum as cheaply as possible? What is the value of the spectrum? 24© Copyright Coleago 2012Spectrum Workshop

Recent auction outcomes show big variations in prices 25 2.6GHz prices paid © Copyright Coleago 2012Spectrum Workshop

Very large variations in prices paid are due to different auction dynamics and levels of competition 26 Country Year €/MHz/pop Commentary Hong Kong 2009 0.252 5 bidders for 3 blocks Sweden 2008 0.130 5 bidders for 4 blocks; one new entrant Denmark 2010 0.164 4 bidders; 2nd price auction, low spectrum caps Norway 2007 0.036 2 operators plus Craig Wireless Germany 2010 0.028 4 operators and 140MHz; substitute spectrum available Finland 2009 0.004 3 operators and 140MHz; single TDD block sold for 50% higher price than FDD Netherlands 2010 0.001 3 operators and two cablecos; low spectrum caps, and 2nd price rule 2.6GHz FDD prices paid in recent auctions: © Copyright Coleago 2012Spectrum Workshop

Digital dividend spectrum is more expensive: Demand and supply is the key driver 27 0.73 0.03 0.11 0.02 0.02 - 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 800 MHz Paired 1.8 GHz Paired 2.0 GHz Paired 2.6 GHz Paired 2.6 GHz Unpaired €/Mhs/PopsGerman Spectrum Auction Prices €/MHz/Pop © Copyright Coleago 2012Spectrum Workshop

Valuing spectrum is conceptually straightforward 28 Net Present Value with NPV without Valuation Business value with extra spectrum Business value without extra spectrum Maximum value of extra spectrum The valuation needs to take mitigating strategies into account: We compare optimal ‘acquire’ with optimal ‘non-acquire’ cases (i.e. assume optimal strategy is pursued in all cases) © Copyright Coleago 2012Spectrum Workshop

Valuations are typically based on a 10 year forecast and a terminal value using Discounted Cash Flow Analysis 29 Free Cash Flow Forecast NPV of 10 Year Forecast Terminal Value EV Valuation Aligned with management forecasts in the early years Valuations are then subject to benchmarking using company comparables such as EV / EBITDA © Copyright Coleago 2012Spectrum Workshop

How value is created from spectrum is not always fully appreciated  There are two options to end an auction:  Option A, you could obtain 20 MHz – You value 20 MHz at € 500 million – You can obtain it for € 480 million  Option B, you could obtain 10 MHz – You value 10 MHz at € 300 million – You can obtain it for € 240 million 30 Which would you choose? © Copyright Coleago 2012Spectrum Workshop

Value creation arises where spectrum valuation and the price paid in the award process meet  There are two options to end an auction:  Option A, you could obtain 20 MHz – You value 20 MHz at € 500 million – You can obtain it for € 480 million  Option B, you could obtain 10 MHz – You value 10 MHz at € 300 million – You can obtain it for € 240 million 31 Value created: € 20 million Value created: € 60 million © Copyright Coleago 2012Spectrum Workshop

It’s about increasing shareholder value, not about acquiring the largest or most valuable spectrum blocks 32 Value placed on a block Price paid Value created  Oddly, the point about value creation is often the most difficult to communicate to senior management. € / $ © Copyright Coleago 2012Spectrum Workshop

33 Increasingly there are multiple-bands to be valued Business value with spectrum Business value without spectrum Maximum values of spectrum © Copyright Coleago 2012Spectrum Workshop

Synergies between and within blocks requires a holistic approach 34 800MHz 900MHz 1800MHz 2100MHz 2600MHz © Copyright Coleago 2012Spectrum Workshop

The degree of complexity and the number of valuations can be difficult to comprehend Spectrum valuation is increasingly complex  Large number of feasible spectrum combinations all require values  Edge-block interference means the position of spectrum blocks is important  Key uncertainties around demand and technology require a scenario-driven approach 35 Band Mode Available MHz 800MHz FDD TDD 2 x 30 900MHz FDD 2 x 35 1800MHz FDD 2 x 75 2.1GHz FDD TDD 2 x 60 1 x 20 1 x 15 2.6GHz FDD TDD 2 x 70 1 x 45 Total 620MHz Example: Swiss Spectrum Auction © Copyright Coleago 2012Spectrum Workshop

Operational sources of spectrum value Blocking value Capacity competition Quality & coverage competition Access speed claim competition Site-build capex avoidance Technology migration benefits Any assumptions that have a market share impact will lead to high spectrum valuations Technically related sources usually generate lower spectrum valuations 36© Copyright Coleago 2012Spectrum Workshop

Non operational sources of spectrum value Tax shield Terminal value assumptions 37© Copyright Coleago 2012Spectrum Workshop

A good general philosophy 38 “It’s better to be roughly right than precisely wrong” John Maynard Keynes © Copyright Coleago 2012Spectrum Workshop

Issues in spectrum valuation Spectrum valuation in practice 39© Copyright Coleago 2012Spectrum Workshop

40 Spectrum valuation must take all potential resources into account, in different parts of the network Deep indoor Deep indoor RuralUrban Feasible coverage with low band spectrum only Addressable with all bands Macro network Macro densification layer Capacity solutions layer Lower scope for cell- splits in (dense) urban given high site density Higher scope for cell- splits in rural given low site density Outdoor: small cells Indoor: Femto cells and WiFi Limited scope for small cells and WiFi offload © Copyright Coleago 2012Spectrum Workshop

Identifying industry bottlenecks is paramount Technology roadmap and device penetration  Low (initial) penetration of LTE-enabled devices limit the utilisable capacity in LTE bands such as 800MHz and 2.6GHz – 3G capacity likely to remain the bottleneck in the near/medium term  The site upgrade path may have a significant impact on the valuation Regulatory network deployment constraints  Severe emission limits may prevent operators from deploying additional mobile channels on individual sites (e.g. Switzerland)  Planning constraints may limit the scope for site build, and constrain site densification (e.g. no permission to build on residential property in Ukraine) The problem may be capacity, in-building quality, geographic coverage, or all three  Securing the right mix of spectrum may be as important as obtaining the right quantity 41© Copyright Coleago 2012Spectrum Workshop

Spectrum valuation must take account of mitigating strategies  Capacity solutions – Indoor: Femto cells, Private WiFi – Outdoor: small cells – Traffic offloading on public WiFi  Yield management / repositioning the offer – Focus on existing and new voice customers (converged offerings)? – Prioritize access to network capacity to high-value customers?  Infrastructure-sharing – Potential source of capacity and coverage synergies as well as cost rationalisation  Future spectrum releases?  Buy capacity from a competitor?  Mitigate limited geographic footprint through national roaming? A valuation must reflect all that is relevant to the asset being valued 42© Copyright Coleago 2012Spectrum Workshop

Challenging strategic questions need to be addressed Should operators deploy HSPA in 900MHz or go straight to LTE?  900MHz enabled HSPA devices widely available today  Potential first mover advantage  LTE offers higher performance  Migrating from to LTE from HSPA in the future introduces extra costs Should operators pursue a strategic or tactical LTE deployment?  Strategic: urban or national LTE footprint? – What is the business case?  Tactical: only deploy LTE as a capacity resource in busiest parts of the network? How much GSM spectrum can or should one re-farm for 3-4G?  When can GSM be switched off? 43© Copyright Coleago 2012Spectrum Workshop

Generic methodological principles Spectrum valuation in practice 44© Copyright Coleago 2012Spectrum Workshop

Key sources of spectrum value 45 Greater capacity (all bands) Fewer sites for given coverage and capacity (all bands) More traffic Greater coverage footprint (low band) Extra Spectrum Higher costs of sale & overheads In-building quality (low band) Greater performance (2x20MHz LTE) More customers Higher network costs Lower network costs Spectrum roll-out costs More revenues © Copyright Coleago 2012Spectrum Workshop

Valuing capacity: Network congestion puts customers and revenues at risk 46 % of cells overloaded Mbps per Cell Network Cells (ranked by peak demand) Maximum Throughput per Cell (including densification uplift and other capacity solutions) Need to take an industry-wide view: consider congestion levels versus any spare capacity across all operators Unserved demand leads to loss of customers and revenues if there’s a better alternative elsewhere Unconstrained Peak Demand per Cell Constrained Peak Traffic per Cell © Copyright Coleago 2012Spectrum Workshop

Illustration  Operator A has more spectrum, leading to higher cell throughput  Operator A is positioned as a quality network, and prices its services at a premium  Operator B is a low cost operator; as a result, customers are more willing to accept a lower quality of service during peak hours  Operator B has a lower cell throughput, but higher overall congestion tolerance Valuing capacity: Customer migration between networks will depend on relative congestion-tolerance levels 47 Cell Throughput (Mbps) Congestion- Tolerance Threshold Cell Throughput (Mbps) Congestion- Tolerance Threshold Operator A Operator B © Copyright Coleago 2012Spectrum Workshop

Valuing capacity: Which customers have highest propensity to churn? The relative impact of congestion across key segments is likely to reflect  Degree of customer exposure to the problem (hence sensitivity to congestion)  Yield management strategies pursued by operators – De-prioritising least desirable customers (lowest % margin contributors) 48 Device type Segment Description Average % margin Sensitivity to congestion Small screen (“HS”) “Standalone HS Low” Low data usage; not sensitive to congestion High (low data) Negligible “Standalone HS High” High small-screen data usage High (limited data) Moderate “Linked HS-PC” Correlated small & large- screen; incl. tethering Medium (heavy data) High Large- screen (“PC”) “Standalone PC” Independent of small screen voice and data Low (heavy data no voice) High © Copyright Coleago 2012Spectrum Workshop

49 Valuing in-building quality: Benefit of Low Band spectrum Probability of service With spectrum below 1GHz With spectrum above 1GHz Deep indoor RuralUrban Deep indoor  In-building quality benefits of LB spectrum can be expressed in terms of addressable demand – Without low band spectrum, none of this ‘deep indoor’ demand can be served – Unless potentially costly indoor solutions are deployed – Such solutions can only solve part of the problem  Cannot be mitigated using yield management strategies  Hong Kong: 850MHz worth 5x more per MHz than 2.6GHz spectrum!  Value of additional blocks of low band spectrum can be quantified using the ‘capacity-value’ approach % Demand © Copyright Coleago 2012Spectrum Workshop

Valuing deep indoor capacity: Additional Low Band spectrum increases the % demand that can be served 50 % of cells overloaded with 1 block Deep indoor Mbps per Cell Network Cells (ranked by peak demand) Max Indoor Throughput per Cell with 2 LB Blocks The first block of low band spectrum provides the largest increase in addressable indoor demand; subsequent blocks generate decreasing marginal returns Constrained Deep Indoor Traffic per Cell Max with 1 LB Block Extra demand addressable with additional LB block Unconstrained Deep Indoor Demand per Cell © Copyright Coleago 2012Spectrum Workshop

 Low Band spectrum may enable an operator to extend the geographic footprint  Potential impact on addressable market, hence market share (if significant rural demand exists)  Impact of wider footprint on marketing productivity – Share of gross ads, retention rate (reduced churn) – Scope for price premium (ARPU impact) 51 Valuing geographic coverage: Low Band impact on footprint Deep indoor Deep indoor RuralUrban © Copyright Coleago 2012Spectrum Workshop

Performance (widest LTE channel)  One 2x10MHz LTE channel provides greater performance for a given level of traffic than two 2x5MHz channels,  One 2x20MHz LTE channel provides greater performance for a given level of traffic than two 2x10MHz channels, but only marginally  Potential impact on marketing productivity of “speed claim” / experienced performance Network cost avoidance  To achieve a given level of capacity: deploying spectrum on existing sites is typically less expensive than splitting cells and deploying micro solutions (indoor: Femto cells; outdoor: small cells)  To achieve a given geographic footprint: rolling out a network with lower bands is less expensive (and yields better quality of coverage) than doing so with higher bands Positioning of spectrum blocks can also impact on value  Edge block interference (e.g. 2.6GHz FDD versus TDD)  Interference with digital terrestrial TV (800MHz) 52 Other spectrum value components © Copyright Coleago 2012Spectrum Workshop

Model structure overview 53 Customers, ARPU/AUPU, Revenues Network Dimensions Network Costs Scenario Manager, User Interface Traffic Segmented Analysis Total Market Demand Capacity / Congestion Analysis Consolidation Model Network ModelMarket Model Market Scenarios Technical Scenarios Regulatory Scenarios Competition Scenarios Coverage Analysis Spectrum valuation: cash flows ‘With’ minus ‘Without’ extra spectrum Cash flows, Financial Statements Device Diffusion Traffic Distribution Profile All MNO’s (Industry- wide view) © Copyright Coleago 2012Spectrum Workshop

Spectrum renewal costs also bear on the terminal value of new spectrum acquired Spectrum Renewal 54© Copyright Coleago 2012Spectrum Workshop

Uncertainties around the conditions, process and costs of spectrum renewal pose further challenges for operators Typical approaches to spectrum renewal  Setting of renewal fees using Administered Incentive Pricing (AIP)  Re-auctioning existing spectrum holdings upon licence expiry (e.g. Singapore 2008 for 900MHz renewal, proposed in Switzerland and the Netherlands  Private, bi-lateral negotiations between regulator and individual operators, with threat of auction if no agreement is reached (e.g. Australia) Typical regulatory objectives  Secure an adequate return to society and incentivise efficient use of spectrum  Maintain competition  By setting prices that will appear: – Comparatively cheap for those who make best economic use of it and comparatively expensive for less efficient users of the spectrum, thus ‘incentivise’ more efficient use of spectrum resources © Copyright Coleago 2012Spectrum Workshop 55

Operators who face licence expiry are usually better off with a negotiated settlement rather than a re-auction  Operators faced with the expiry in their licence and refarming usually go through a negotiation process  Case study: Australian 800MHz, 1.8GHz, 1.9GHz TDD and 2.1GHz renewal – Separate negotiations with each operator on renewal fees – Threat of auctioning the spectrum to be renewed if individual operators fail to reach negotiated agreement  Developing a compelling negotiation case for favourable spectrum renewal terms is essential – Covering both renewal fees as well as conditions (e.g. technology neutrality) – Ensuring competitors may no less per MHz, and ideally more!  “Independent” reports commissioned by operators that support their case may help to sway the argument in their favour 56© Copyright Coleago 2012Spectrum Workshop

Administered Incentive Pricing (AIP)  Most regulators are obliged to agree an allocation methodology for renewal that: – yields a return for society and ensures that economic benefit is generated – ensures that the mobile market is sufficiently competitive  Recognising that a simple re-auction may not be appropriate, many regulators have opted for negotiated renewal prices using different methods, such as: – Spectrum fee benchmarking – Optimum Deprival Value (ODV) – Best Alternative Use (BAU) 57© Copyright Coleago 2012Spectrum Workshop

Benchmarking  Benchmarking is the process of setting prices against a proxy taken from another, comparable market.  Benchmarks do not reflect valuations but prices paid.  International benchmarks suffer problems of translation to the local context. – Spectrum demand and supply are very different is each country – Markets vary in terms of population, population density, geography, income levels / willingness to pay and input cost levels (e.g. staff costs). – Technically, benchmarking is further affected by such factors as choice of exchange rate, the comparator set and so on. – Timing also matters, since perceptions of the prospective value of mobile spectrum has varied significantly over time.  Benchmarks are not considered as a suitable method in setting spectrum prices, but cash strapped governments have used them to justify high reserve prices. – A situation best avoided, but depends on regulatory policy © Copyright Coleago 2012Spectrum Workshop 58

Optimal Deprival Value (ODV) – Academic description Deprival value is a cost-based valuation approach that answers the question:  “What is the least cost system or bundle of assets needed to provide customers with the existing level and quality of services, should a certain existing asset be removed?” This approach can be applied to valuing spectrum rights by addressing the following:  “If an operator was deprived of incremental spectrum rights, what incremental costs would be incurred to replicate the existing level and quantity of services using the remaining spectrum rights?”  These costs are “avoided” by owning the incremental spectrum rights and so represent the value of those rights. The rights holder should be prepared to pay up to the value of the incremental costs to avoid being deprived of its spectrum rights, so long as the incremental costs are less than the present value of the free cash flows generated from the spectrum services. 59© Copyright Coleago 2012Spectrum Workshop

Optimal Deprival Value (ODV)  ODV calculates this price on the basis of the ‘deprival value’ of a marginal block of spectrum (typically 5MHz), for an ‘average reference operator’ – Deprival value: the network costs that would need to be incurred to offer the same quality of mobile service, if the marginal spectrum block were lost – The only source of value is technical, hence ODV produces a low value  Since this is calculated for an ‘average operator’ with ‘average spectrum holdings’, the ODV price per MHz should appear: – Less expensive for operators with less spectrum, who would value it more highly – More expensive for those with surplus spectrum, who should value the marginal block less highly – In this sense, ODV would reflect an ‘optimal price’ in terms of incentives, by encouraging ‘spectrum hoarders’ to give up marginal blocks to more efficient users  ODV generates a uniform price per MHz in any band, for all operators © Copyright Coleago 2012Spectrum Workshop 60

Best Alternative Use (BAU)  In contrast to ODV, BAU considers the value placed on an incremental block of spectrum by alternative users or uses, such as: – An alternative technological use – A new mobile entrant – Competitors  The case for new mobile market entry is typically weak, and the alternative technological uses of mobile spectrum are invariably less valuable than mobile  Hence BAU is normally driven by the value placed by existing competitors on a marginal block of spectrum – Unlike ODV, BAU may generate different prices per MHz for each operator  If all other things are equal, BAU should generate a higher price per MHz for operators with spectrum surpluses, and a lower price per MHz for those with smaller holdings © Copyright Coleago 2012Spectrum Workshop 61

In some markets, regulators propose to align renewal fees of expiring licences with auction outcomes for new spectrum  Current consultation in the UK: base 900MHz and 1800MHz renewal fees on forthcoming 800MHz and 2.6GHz auction – 900MHz renewal based on 800MHz proceeds – 1800MHz renewal based on linear average between 800MHz and 2.6GHz proceeds (!)  India: proposal to base 1800MHz renewal on proceeds of 2.1GHz auction  No longer an auction for new spectrum, but a simultaneous auction and renewal fee negotiation! © Copyright Coleago 2012Spectrum Workshop 62

How to obtain the spectrum as cheaply as possible Auction strategy 63© Copyright Coleago 2012Spectrum Workshop

No auction design is perfect © Copyright Coleago 2012Spectrum Workshop 64 Design Principle SMRA SMRA-AS CCA Supports simultaneous award of spectrum in multi-bands ✔✔✔ ✔✔✔ ✔✔✔ Reduces exposure or aggregation risk ✘ ✔ ✔✔✔ Flexibility over the use of specific or generic lots ✔✔ ✔✔ ✔ Transparency of bidders and bids ✔✔✔ ✔✔✔ ✔✔ Certainty over prices paid ✔✔✔ ✔✔✔ ✘ Certainty over lots awarded ✔✔✔ ✔✔✔ ✘ Certainty over total expenditure ✔✔✔ ✔✔ ✘ Simplicity and ease of presentation and transparency of results ✔✔✔ ✔✔✔ ✘

CCA second price auctions Key characteristics  Winners do not pay the price they bid but a function of the next best allocation of resources (if the winning bidder were absent) – Very complex combinatorial algorithms – Predicting prices paid is challenging for bidders  The price paid by any bidder is determined by marginal valuations expressed by competing bidders  Prices paid per lot can differ significantly between bidders – It is possible for a bidder to pay more for less spectrum than a competitor! 65© Copyright Coleago 2012Spectrum Workshop

Real Example: impact of second price rule in the Denmark CCA based 2.6GHz auction in 2010  Hutchison paid €0.9 million for 2x10 MHz of FDD plus 25 MHz of TDD  The other bidders who acquired 2x20 MHz FDD paid ~20x more per MHz  This dramatic outcome was a product of a second price combinatorial auction with tight spectrum caps: – TDC, Telenor and Telia’s prices reflected Hutchison’s bid value for an additional lot of 2x10MHz FDD – Hutchison’s 2x10MHz FDD could not have been sold to anyone else, hence the 2nd price was the reserve price 66 *Prices paid per MHz © Copyright Coleago 2012Spectrum Workshop H3G TDC Telia Telenor

Conclusions 67 Spectrum auctions are increasingly complex  These are multi-dimensional problems  Complex game-theoretical aspects  A thorough understanding of the auction rules and the risks and opportunities is critical Valuations for each feasible package is required  Not just own valuations, but also the (likely) valuations of other bidders  Extensive competitor analysis and intelligence gathering is required (cash position, strategic objectives, public announcements, past auction behaviour,…) Implementation work-streams need to start well in advance  Influence auction rules at consultation stage  Influence government spectrum policies © Copyright Coleago 2012Spectrum Workshop

Overview of auction strategy development and auction readiness © Copyright Coleago 2012Spectrum Workshop 68 1 Auction objectives, valuations, bid limits Understand the work of the valuation team. To agree auction objectives. Set bid limits and explain rationale. 2 Develop competitor intelligence Identify competing bidders and ascertain their auction objectives, auction strategies and bid limits relative to own valuation. 3 Develop bid strategy Generate actionable insight into the auction format and rules, and into the underlying opportunities, risks, and mitigating strategies. 4 Develop bid optimisation & support tools Create the tools that allow Coleago to validate bid strategy and support the implementation of the bid strategy in a live auction environment. 5 Bid strategy testing and refinement Validate strategic hypotheses through auction simulations and quantify risks and opportunities 6 Mock auctions & auction readiness Test the bidding strategy and to train the auction team and to test auction support tools, bid room protocols, procedures and disaster recovery as well as providing insight to senior management . 7 Game plan To be able to execute the bid strategy in accordance with robust corporate governance.

Understand valuations, bid limits and auction objectives as well as highlighting auction risks An initial workshop with senior management is often useful to highlight the key issues that will arise in the auction A typical workshop agenda is as follows  Overview of the spectrum on offer and any caps  High level overview of the auction process  Explanation of key auction issues and risks that must be addressed  Discussion and agreement of primary and secondary auction objectives  Explanation of the role of bid limits and agreement on the nature of any limits to be imposed on the bid team  Discussion surrounding management’s attitude towards risk taking in the auction Spectrum Workshop © Copyright Coleago 2012 69

Competitor intelligence Competitor intelligence is critical in this auction  The valuation workstream also provides valuation matrices for all other bidders  We will need to make an assessment of other bidder’s likely bid limits  We analyse other bidders relative to your operation and adjust their bid limits accordingly  The exercise is highly subjective so having a framework is useful Bidder Limit Analysis Bidder 1 Bidder 2 Bidder 3 Bidder 4 Financial Strength 1 1 3 3 Strategic Imperative 2 3 2 2 Previous Bidding History 3 2 1 3 Public Announce- ments 1 3 3 2 Capacity Constraints 2 3 1 1 Spectrum Workshop © Copyright Coleago 2012 70

Develop bidding strategy In the many auctions there is no “dominant strategy”  The academics will analyse the auction rules in detail in conjunction with – Your auction objectives, valuations and budget limits – Competitor intelligence, valuations and bid limits  The analysis will allow them to develop a range of potential auction strategies  We will then “War Game” the potential strategies to – Test how robust each strategy is to different competitor strategies – Identify the actions of other bidders that would cause a change in strategy – Develop an initial “Auction Game Plan”  We would present the initial “Game Plan” at a bid strategy workshop Spectrum Workshop © Copyright Coleago 2012 71

Develop the auction simulation tool Spectrum Workshop © Copyright Coleago 2012 72

Identify and develop appropriate live auction support tools With potentially only ten minutes per round many aspects of bid execution will need to be automated There are a range of auction support tools that may be appropriate  Auction bid tracker – Tracks historic bids (audit and governance) – Predicts when bid limits may be breached for both you operation and other bidders based on bid increments and bid levels – Analyses dynamic lot valuations and prices and provides data on absolute and relative value creation  Bid optimisation tool – Utilises the results from previous auctions and updates spectrum values for current and future auctions and generates automated recommendations for bids Spectrum Workshop © Copyright Coleago 2012 73

An example of a bid optimiser from the Canadian AWS spectrum auction Spectrum Workshop © Copyright Coleago 2012 74

Bid strategy testing and refinement  Extensive auction simulations will be used to test the impact of deviations from the ‘dominant bid strategy’ under numerous competitive scenarios, covering: – Alternative competitor valuation and cash-constraint assumptions – Alternative ‘price discovery’/ bid inflation assumptions – Individual and collective deviations from ‘sincere bidding’ strategies including strategic demand reduction and ‘eligibility parking’  Key outputs – Comprehensive testing of alternative bid strategies under a broad range of assumptions to validate approach – Identification and quantification of risks and opportunities – Empirical validation of supplementary-bid optimisation and price prediction capabilities (subject to auction format) © Copyright Coleago 2012Spectrum Workshop 75

Mock auctions – bid team training Mock auctions are often more about training the bid team than developing bid strategy We recommend running a minimum of two mock auctions  Plan the mock auction process and schedule of auctions  Identify and invite the right participants  Prepare instructions and bidder briefs, including bid limits for each mock auction for all participants  Explain the auction design, key issues and mock auction process  Run the mock auctions, providing support and training as required The results and conclusions will be presented at a final auction bid strategy workshop where the strategy will be refined if appropriate Spectrum Workshop © Copyright Coleago 2012 76

Mock auctions – testing protocols and procedures The Belgian auction will require extremely robust auction room protocols and procedures The mock auctions provide an opportunity to test all aspects of bid strategy execution  The effectiveness of the live auction support tools  Bidding protocols – Deciding the bid strategy, preparing the next bid, entering the bid, checking the bid, submitting the bid, updating the auction activity journal  Escalation procedures in the event of anticipated bid limit breaches  Disaster recovery Spectrum Workshop © Copyright Coleago 2012 77

Auction Game Plan Document  Pre-auction activity – e.g. aggressive statements prior to the auction  Statement of hierarchy of primary and secondary auction objectives  Your valuation and bid limit matrices  Competitor bid limits and triggers for re-assessment of competitors  Opening auction bid strategy  Description of bidding patterns which might indicate strategic behaviour  Description of trigger points for switching auction objectives / strategy  Alternative bidding strategies in response to strategic behaviour  Triggers for bid limit escalation procedures  Bid team roles and responsibilities  Bidding protocols and procedures  Disaster recovery  Post auction press releases prepared for win / lose outcomes Spectrum Workshop © Copyright Coleago 2012 78

Summary 79© Copyright Coleago 2012Spectrum Workshop

Summary  Operators typically under estimate the complexities and challenges associated with spectrum valuation – Often regarding auction bidding strategy as the more challenging  Early engagement with regulators is essential to ensure that any spectrum award or renewal process is favourable to you  Coleago can support you across a broad range of issues – Regulatory strategy and lobbying – Support during formal spectrum consultation – AIP lobbying and modelling – And of course, spectrum valuation, auction strategy and readiness and live auction support 80© Copyright Coleago 2012Spectrum Workshop

Steps in Network Consultation APPENDIX 1 81 Appendix © Copyright Coleago 2012Spectrum Workshop

Typical steps in a public spectrum consultation process © Copyright Coleago 2012Spectrum Workshop 82 Stage Activity Operator’s Actions Policy Formulation The Government formulates policy for the allocation of spectrum. Government departments or ministries involved may include Telecoms, Development, Finance, Competition Commission, etc Lobby politicians PR campaign to influence the debate Notice of Consultation At this stage the process moves to the telecoms regulatory authority who issues a notice of consultation. This includes a description of the consultation process and timetable. There may also be a workshop to obtain inputs from stakeholders in an informal manner. Lobby politicians PR campaign to influence the debate Prepare workshop materials Meetings with regulatory agency 1st Round Consultation The regulator issues a formal consultation document which contains a list of decisions that need to be made in relation to the spectrum allocation, a discussion and analysis of the issues, perhaps a preliminary conclusion (“we are minded to....”), and an invitation to respond to questions relating to each decision that needs to made. Analyse consultation document Commission independent reports that support own view point Prepare & submit response to consultation PR campaign to influence the debate Cross Submission The regulator would normally publish the submissions received from different stakeholders and invite cross submissions. This gives stakeholders the opportunity to comment on points made by other stakeholders. Analyse other submissions quantitatively (how many agree with x) and qualitatively Prepare and file cross submission document PR campaign to influence the debate 2nd Round Consultation The regulator issues a second consultation document which draws on the submission received and presents its conclusion and a proposal ahead of the final determination. If an auction is proposed this would include a description of the auction format and rules. Analyse consultation document Submit response to consultation PR campaign to influence the debate Final Determination The regulator publishes the final determination, including the process, detailed procedures and rules. The regulator answers questions for clarification of the process, procedures and rules. Analyse determination document Questions for clarification Explore legal recourse if necessary

Approach to modelling APPENDIX 1 83 Appendix © Copyright Coleago 2012Spectrum Workshop

Approach to Network Modelling Overall approach: Options:  Calculate traffic offloaded onto public WiFi and/or small sites, subject to separate offload limits in different parts of the network; model using abstract cost per Mbps  Treat densification as a capacity resource relative to existing sites, subject to separate densification limits in different parts of the network 84 Compare traffic in each 1% of sites (all operators) Adjust traffic on basis of relative congestion (all operators) Calc. resources required to meet adjusted traffic (single operator)Compare max capacity in each 1% of sites (all operators) Calculate capex and opex (units x unit costs) - Spectrum - Technologies - Densification, etc. - RAN - Backhaul - Core LTE roll-out (single operator) Appendix © Copyright Coleago 2012Spectrum Workshop

Model structure overview 85 Customers, ARPU/AUPU, Revenues Network Dimensions Network Costs Scenario Manager, User Interface Traffic Segmented Analysis Total Market Demand Capacity / Congestion Analysis Consolidation Model Network ModelMarket Model Market Scenarios Technical Scenarios Regulatory Scenarios Competition Scenarios Coverage Analysis Spectrum valuation: cash flows ‘With’ minus ‘Without’ extra spectrum Cash flows, Financial Statements Device Diffusion Traffic Distribution Profile All MNO’s (Industry- wide view) Appendix © Copyright Coleago 2012Spectrum Workshop

Impact of spectrum scenario on Reference Demand 86 Customers ARPU and AUPU Reference Demand Revenues and Traffic Customers ARPU and AUPU Pre-Congestion Demand Revenues and Traffic Customers ARPU and AUPU Congestion- Adjusted Demand Revenues and Traffic Adjustments based on all commercial value drivers except capacity: – Relative footprint – Quality – Performance Impact on: – Share of gross ads – Churn – ARPU Adjustments based on relative congestion: – Equalisation of congestion levels – Subject to relative congestion tolerance limits – Traffic and subs migrate from more to less congested networks Impact on: – Share of gross ads – Churn Traffic Distribution Profile Capacity / Congestion Analysis Net Traffic Migration Within Consolidation Model Within Technical Model Appendix © Copyright Coleago 2012Spectrum Workshop

Valuing capacity: Combined congestion-impact biases 87 “Standalone HS High” 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Relativetoaverage % of base Relative AUPU versus ARPU ARPU AUPU “Linked HS-PC” 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Relativetoaverage % of base Relative AUPU versus ARPU ARPU AUPU “Standalone PC” 0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 0.0% 20.0% 40.0% 60.0% 80.0% 100.0% Relativetoaverage % of base Relative AUPU versus ARPU ARPU AUPU Key segments: Bias within each segment: Bias across segments: Highest (e.g. 60%) Next highest (e.g. 25%) Lowest (e.g. 15%) Net impact of congestion: High impact on traffic relative to revenues Moderate impact on traffic relative to revenues Low impact on traffic relative to revenues Appendix © Copyright Coleago 2012Spectrum Workshop

Modelling workflow (overview) 88 Reference Demand (all operators) Within Consolidation Model Pre-Congestion- Demand (all operators) Spectrum impact excl. capacity Congestion Impact Congestion- Adjusted Demand (all operators) Impact on % congestion tolerance (all operators) Peak traffic across network (all operators) Congestion analysis (all operators) Within Network Model Net traffic migration (all operators) Adjusted traffic (all operators) Capacity upgrade requirements (selected operator) P&L and Cash Flows (selected operator) Network dimensions (selected operator) Unit costs Capex & Network Opex (selected operator) Link budgets, 3G+ throughputs Coverage analysis (all operators) Max effective capacity per cell (all operators) Costs of Sale & Overheads (selected operator) Appendix © Copyright Coleago 2012Spectrum Workshop

Add a comment

Related presentations

Related pages

Workshop on Market Mechanisms for Spectrum Management

... management and related spectrum issues. ... Workshop on Market Mechanisms for Spectrum ... Spectrum Valuation and Spectrum ...
Read more

Communication Workshop: Issues and challenges in decision ...

Spectrum Workshop - Issues and challenges in spectrum allocations and spectrum valuation
Read more

Valuing Spectrum Allocations (PDF Download Available)

... Valuing Spectrum Allocations on ... deem the Broadcast TV Spectrum Valuation . ... to summarize the challenges facing spectrum ...
Read more

OPM model and spectrum - Government & Nonprofit

Spectrum Workshop - Issues and challenges in spectrum allocations and spectrum valuation
Read more

Spectrum Management Overview | www.ictregulationtoolkit.org

This is known as making spectrum allocations on ... Valuation. Digital Terrestrial ... In the administrative spectrum management method, a spectrum manager ...
Read more

National Broadband Plan - Chapter 5: Spectrum

... a triennial assessment of spectrum allocations. ... average spectrum valuation for mobile broadband use of $1 ... FCC Spectrum Workshop ...
Read more

Spectrum Valuation | LinkedIn

Current Spectrum Valuation Analyst at Telefónica, CapEx Investment Analyst at Telefónica Past Junior Corporate Strategy Analyst at Telefónica UK ...
Read more

Spectrum value - ScienceDirect

... DCF and related cost savings analyses are useful for valuing new or proposed spectrum allocations, ... Spectrum valuation ... Workshop on Spectrum ...
Read more