The Delta Perspective

April 2011

The Digital Dividend in Europe: in the eye of The Perfect Storm

Authors Alberto Pamias - Managing Partner Daniel Torras - Associate Partner

A USD 50 Bln rule-changing play triggered by the new 800 MHz spectrum. In this high-stakes game old rules will no longer apply and operators will need to reinvent themselves to survive.

OVERVIEW

The 800 MHz frequencies, commonly referred to as the Digital Dividend, have the potential to unleash a perfect storm in the telecommunications industry. Compounding the need for operators and vendors to capitalise on the growing demand for mobile broadband, the scarcity of available spectrum and the governments opportunity to cash in billions of dollars in spectrum auction proceeds, the Digital Dividend may disrupt the global competitive landscape in an unprecedented manner and force operators to make a wide range of strategic decisions of significant impact, including revisiting their strategic positioning and commercial strategies, partnering with competitors, reevaluating technology choices and rolling out new technologies and networks, refarming spectrum if necessary. This white paper explores the significance and implications of the Digital Dividend and concludes with a strong call to action for operators to evaluate all options and develop an action plan now. The complexity and breadth of decisions that CEOs of European operators need to make, and the imminent award of new licences across the continent, require that this be tackled as a matter of urgency.

The Digital Dividend in Europe: in the eye of The Perfect Storm

In meteorological terms, a storm is a disturbance of the atmosphere marked by wind and usually by rain, snow, hail, sleet, or thunder and lightning1. A severe storm can have a negative impact on peoples lives and property. A perfect storm is a rare combination of circumstances that aggravates the situation drastically.
The global telecommunications industry, and especially the mobile sector in most of the developed and an increasing number of the developing markets, is sailing into a perfect storm. Following the meteorological analogy, three different weather-related phenomena are coalescing in the formation of this telecommunications storm: Warm air from a low-pressure system coming from one direction: the rise of wireless broadband The surge in mobile data as a result of the advent of data-centric handsets, spurred by Apples iPhone and iPad and later followed by the whole Industry, has dramatically changed the way people use the Internet on their mobile devices. Global mobile data traffic is expected to increase 26-fold in the next 5 years to reach over 6 Exabytes by 2015 (from 0.24 Exabytes in 2010)2 A flow of cool and dry air generated by a high pressure front from another direction: the operators need for new sources of growth as markets stagnate Mobile operators are looking for new growth avenues to counter the stagnation of traditional voice services. The data opportunity is the new El Dorado. Pressed to deliver returns to their shareholders, operators are making significant Capex investments to upgrade networks from UMTS to HSPA+, and more recently to LTE, on the promise of a new and long life for the wireless industry. Tropical moisture: the need for additional spectrum and new technological standards to cope with the exponential increase in data traffic Network utilisation is on the rise as a result of the exponential growth in data traffic, leading to severe network congestion for many operators. Operators desperately need additional spectrum and new spectrum-efficient technology bearers in order to sustain data growth. From a policy and regulatory point of view, increasing broadband availability is, on paper, a key priority for governments (in reality, governments motivations can be quite different, as we well see later). In its Digital Agenda published in May 2010, the European Commission stated its goal that member states should provide basic broadband connectivity to 100% of Europeans by 2013, fast broadband (above 30 Mbps) to 100% of Europeans by 2020, and ultra-fast broadband (above 100 Mbps) to at least 50% of European households by 2020.

1 2

Merriam Webster dictionary Cisco

Similarly to a perfect storm, aggravating circumstances are coming together, which are drastically contributing to make this telecommunications storm a perfect one: Device manufacturers are focusing on data hungry devices such as smartphones and tablets to foster (and capitalise on) growth opportunities, pushing infrastructure vendors to engage in an arms race to deliver the best, fastest and cheapest next generation equipment Content providers and consumer advocates have in recent years escalated the debate in favour of net neutrality and closing the digital divide, adding further pressure on operators to improve network coverage and data speeds Many GSM licences that were originally awarded in the 90s are approaching their expiration dates Governments, still trying to recover from the global recession of the 2008-2009 period, need to generate new sources of income to finance their budgets, and the auctioning of scarce public assets such as spectrum has the potential to rake in large sums of money

In the eye of The Perfect Storm is the Digital Dividend, a set of radio frequencies that are very well suited for the provision of mobile broadband, and that have the potential to trigger an unprecedented shake-up in the global telecommunications market. This white paper discusses the business opportunities that the Digital Dividend will unveil (as well as the devastating effects it can have if not implemented well) and how the different market constituents particularly operators - need to prepare themselves, including the trade-offs they must be willing to make, to sail through this perfect storm.

EXHIBIT 1: THE DIGITAL DIVIDEND IN THE EYE OF THE PERFECT STORM

Source: Delta Partners

What is the Digital Dividend and why does it matter?

As the terrestrial television switchover from analogue to digital takes place across the world, some of the radio spectrum that is currently used for TV broadcasting will be freed up and allocated to other uses, such as mobile broadband. This range of frequencies is commonly referred to as the Digital Dividend. A few countries in Europe and the Americas have already completed the switchover, and many others will follow in the next 12-24 months.
In order to facilitate cross-border provision of services, avoid interference and allow equipment manufacturers and operators to benefit from economies of scale and faster time to market, Digital Dividend frequencies have been harmonised globally and divided into 2 parts: one in the lower part of the UHF spectrum band (also known as lower Digital Dividend) and another one immediately above it (upper Digital Dividend)3. The first has been reserved for TV broadcasting, and the second has been reserved for mobile broadband. In Europe and the CIS, as well as Africa, the Middle East and parts of Asia, the band comprised between 790 and 862 MHz will be allocated to mobile broadband.
EXHIBIT 2: ANALOGUE TO DIGITAL TV SWITCOVER TIMEFRAME

Source: GSA

Ultra High Frequency Band: 300 MHz to 3000 MHz

EXHIBIT 3: UPPER DIGITAL DIVIDEND ALLOCATION BY REGION

Source: GSA

From a technical perspective, the Digital Dividend spectrum is very attractive. It is in the lower UHF range, and thus has very good propagation characteristics. Radio signals in this band can travel further distances than higher frequency signals, and are less attenuated by buildings and other geographic or climatologic obstacles. Because a single site operating in the 800 MHz band can cover as much area as up to 5 sites operating in the 2100 MHz band, it is significantly cheaper to provide mobile broadband coverage over a given geographic area using 800 MHz spectrum than with the 2100 MHz spectrum. This advantage is even more pronounced when one compares the Digital Dividend spectrum with the 2600 and 3500 MHz bands, which are also used for wireless broadband. This makes the Digital Dividend spectrum particularly well suited to provide mobile broadband coverage in rural and suburban areas. Another benefit of the Digital Dividend frequencies is that they are very close to the GSM 900 band. Thus, a Digital Dividend base station can in most cases be co-located on existing GSM towers to reuse existing passive infrastructure. Because of these reasons, the Digital Dividend spectrum is highly coveted. As a point of reference, in the May 2010 German auction - which included Digital Dividend spectrum alongside other frequencies in the 1800 MHz, 2100 MHz and 3500 MHz bands - the Digital Dividend frequencies were valued 7 and 32 times more than the 2100 MHz and 2600 MHz spectrum (see Exhibit 4), respectively, attaining a valuation of USD 1.00/POP/MHz.

EXHIBIT 4: PRICES PAID IN GERMANYS MAY 2010 BIG BANG AUCTION (USD CENTS/POP/MHZ)

Source: KB Spectrum, Delta Partners

Taking the German valuation as a reference, the Digital Dividend spectrum could be sold for as much as USD 50 Bln once all the countries in Europe (including transcontinental countries Russia and Turkey) have auctioned it off. Beyond the intrinsic benefits of allocating badly needed spectrum, the Digital Dividend also has the potential to dramatically shake up the competitive landscape by giving challengers powerful weapons with which to fight incumbent operators, and it can also be a catalyst for fixed-tomobile broadband substitution. As such, the Digital Dividend may not be a win-win for all market constituents, and operators need to carefully analyse the implications of the Digital Dividend in their markets in order to design an appropriate defence or attack plan, depending on their competitive position. Given the above considerations, it is easy to see how the Digital Dividend can unleash the previously mentioned perfect storm. It reunites all the characteristics required to boost mobile data growth (the basic requirements for a conventional storm), but it comes a time when all the aggravating factors that can turn it into a perfect storm are flowing together. Of particular importance is the role that regulators will play. Notwithstanding the rhetoric of most governments around social development and the need to provide affordable broadband access to the general population, the truth is that when it comes to spectrum allocation, most governments are moved primarily by greed. In trying to maximise their cash-in at all costs, governments may, perhaps unwillingly, unleash The Perfect Storm whose negative consequences can last for a very long time.

The regulators perspective

As directors of the whole process, regulators exert enormous influence in how the Digital Dividend implementation will take place. Is greed good?4
Telecom regulators need to reconcile two seemingly opposed objectives when awarding telecom licences, especially when access to scarce resources such as spectrum is concerned: The social aspect: increasing competition and availability of services for the general population, to benefit society at large, is at the core of this objective. This can be achieved, for example, by establishing licence obligations that require significant investment and job creation The financial aspect: simply put, this entails maximising the revenue windfall for the government from the auctioning of spectrum

To achieve the first objective (the social one), regulators typically resort to coverage and service level requirements with any new licence, and define auction rules to eliminate the possibility of a single operator monopolising the entire spectrum available, via spectrum caps, spectrum trading restrictions, or a minimum number of licences. In the case of the Digital Dividend, coverage requirements in rural areas often constitute a key component of licence obligations. To achieve the second objective (the financial one), regulators are predominantly opting for auctions as the vehicle to award spectrum, in a clear attempt to maximise financial proceeds. Although advocates of auctions will argue that an auction tends to command the fairest value for the spectrum and that it is the most transparent method to award a licence, history shows that auction
EXHIBIT 5: EVOLUTION OF PRICES PAID FOR 3G SPECTRUM IN EUROPE (USD CENTS/POP/MHZ)

Source: Global Spectrum Database, country regulators, Delta Partners (does not include beauty contests) Wall Street. Oliver Stone. 1987. Gordon Gekko full quote: Greed is good. Greed is right. Greed works. Greed clarifies, cuts through, and captures, the essence of the evolutionary spirit
4

processes in the past decade have led to vastly different (and increasingly low as a function of the auction timing) valuations on a comparable basis (USD/POP/MHz). These differences in valuation cannot be explained by the macroeconomic differences among countries alone, and disprove the hypothesis that an auction is the guardian of fair value. On the contrary, they suggest that the timing of an auction is a far more important factor when it comes to spectrum valuation than the award process itself. The early European 3G auctions commanded much higher prices than latter 3G auctions, as is shown in Exhibit 5. Unsurprisingly, regulators in many countries are rushing to auction the Digital Diligence spectrum ahead of other countries in order to take advantage of this early euphoria effect. Another aspect to consider for regulators looking to maximise the revenue windfall, and of great implications for operators, is the expiration dates of the existing GSM licences. Many of the licences that were awarded 15-20 years ago are coming or will soon come to their expiration, and regulators may see this as an opportunity to auction multiple frequency bands at once, similar to what the German regulator did last May, or what the Spanish regulator is planning to do in the coming months. This represents an attractive opportunity for operators that do not have (or have very little) spectrum in the sub-1 GHz band, and conversely, it is a threat for incumbent operators hoarding large blocks of GSM 900 frequencies. Multiple-band auctions further illustrate the regulators ambitions to maximise their cash-in, by using the Digital Dividend as a trigger of a major rule-changing event. Last, but not least, specific auction design parameters, including minimum bidder activity requirements to remain eligible for subsequent auction rounds, and minimum bid increments, can turn a standard auction into an inflationary one.
EXHIBIT 6: EXPIRATION DATES OF EUROPES GSM LICENSES

Source: Global Spectrum Database, country regulators, Delta Partners

In the past, not all the regulators had placed equal emphasis on the social and financial objectives. Nordic countries, for instance, have traditionally focused more on the social benefits of the licences than on increasing governments revenues, and as such they have historically favoured beauty contests over auctions. In Central Europe, the reverse is true. However, in the midst of economic turmoil or for other reasons, countries are changing their historical stance on this issue. Sweden provides a good example. It awarded 3G licences via a beauty contest back in 2000, but switched to an auction in early 2011 to award the Digital Dividend spectrum, without any coverage requirements whatsoever for six of the five spectrum blocks that were auctioned. 8

The two Digital Dividend auctions completed in Europe to date provide a good illustration of the auction mechanics used by two countries that had historically pursued almost polar opposite objectives (i.e. social-driven Sweden and financially-driven Germany). The spectrum valuations attained in both countries should give a good indication of the upper and lower brackets we should come to expect in the upcoming European auctions: in Germany, the Digital Dividend spectrum was sold for USD 1.00/POP/MHz, whereas in Sweden the valuation was USD 0.57/POP/ MHz. In both cases, six blocks of 2x5 MHz were allocated, with a spectrum cap of two blocks per bidder (which led in both Germany and Sweden to three operators acquiring 2x10 MHz of spectrum each). While on paper the main difference between the licences auctioned in the two countries had to do with coverage requirements (in the case of Germany, all licensees were subject to coverage requirements, whereas in Sweden, coverage obligations were attached only to one of the six 2x5 MHz blocks), in the end, other factors, mainly related to the rules of the game including the mechanics and the timing of the auction itself, and the characteristics of the domestic markets and operators, made one process more inflationary than the other.

EXHIBIT 7: SUMMARY OF THE GERMAN AND SWEDISH DIGITAL DIVIDEND AUCTIONS

Source: PTS, BNetzA, Delta Partners

One of the next countries in the line to award the Digital Dividend spectrum is Spain. It is a remarkable case in that Spain has traditionally favoured beauty contests to award spectrum, but has already announced it will use an auction format to award the Digital Dividend spectrum, and there has been no mention of any coverage requirements. The Spanish Government recently published the mechanics it intends to use in its upcoming spectrum allocation process, tentatively scheduled for the second quarter of 2011. Being under a big budget pressure, it is not surprising that several aspects of the Spanish process are blatantly designed to maximise the financial proceeds for the government, even if the Spanish administration purports to disguise this auction as a process aimed to advance social goals. Some of the Spanish auction design elements are quite different from previous auctions in the continent and might set a benchmark going forward:

Spectrum to be awarded: 310 MHz in total, comprising 4 different bands (800 MHz, 900 MHz, 1800 MHz and 2600 MHz); this represents a 70% increase in the total amount of spectrum currently used for the provision of mobile services in Spain Geographic scope: all licences will be national, with the exception of several blocks in the 2600 MHz band, which will be awarded on an regional basis Eligible bidders: Telefonica and Vodafone will not be allowed to bid for the new 900 MHz and 1800 MHz spectrum, and Orange will not allowed to bid for the new 1800 MHz spectrum; other than that, any party can bid for any of the licences Technology neutrality: all licences will be technology neutral, meaning licence holders will be free to use any bearer technology they choose, be it GSM, UMTS, HSPA+, LTE or other future technologies Award format: frequencies will be awarded using both auctions and beauty contests. For the 800 MHz and 2600 MHz bands, auctions will be used. For the 1800 MHz band, there will be a beauty contest. Finally, for the 900 MHz band, a hybrid model combining an auction and beauty contest depending on the spectrum block will be used Fungible spectrum caps: a combined spectrum cap of 2x20 MHz per operator will apply to the 800 MHz and 900 MHz bands, and a combined cap of 115 MHz (unpaired) will apply to the 1800 MHz, 2100 MHz and 2600 MHz bands Flexible spectrum trading: operators will be allowed to trade spectrum starting 2 years after the concession period begins, provided that licence obligations and spectrum cap restrictions remain in force

The reserve price for each of the six 2x5 MHz blocks of Digital Dividend spectrum to be auctioned in Spain has been set to Euro 170 Mln, representing a total minimum valuation for the 60 MHz available of over Euro 1 Bln (around USD 1.4 Bln). This is equivalent to a valuation of just over USD 0.50/POP/MHz. While this appears to be broadly in line with the German and Swedish cases, it is just the reserve price and there is a possibility that the final valuation might be materially higher once the auction is completed. As a point of comparison, the final price paid in Sweden for the Digital Dividend spectrum was more than two times higher than the initial reserve price. The whole telecommunications sector is expectant about the outcome of the Spanish process to understand whether it may become a reference point (and trigger) for a widespread explosion of inflationary auctions throughout Europe. In summary, notwithstanding the importance of spectrum auction proceeds in terms of balancing government budgets, telecom regulators must carefully assess the impact that these auctioning processes can have on their markets. An unwelcome side effect of inflationary auctions might be that some of the existing operators, typically the smaller ones that are often the ones with the greatest need for spectrum, leave auctions empty-handed (as was the case with E-Plus in Germany), while the auction winners are forced to take on massive amounts of debt to pay for the licence fees and network rollout, potentially at the expense of limiting geographic coverage, increasing broadband prices, or downsizing their workforces. In this context, regulators must understand the motivations of operators and be prepared to deal with their lobbying efforts, and design and implement an auction process that best meets the social and financial objectives defined earlier.

10

The operators perspective

1. The complexity of decisions involving the Digital Dividend
The decision of whether to acquire Digital Dividend frequencies or not is a very complex one, and goes far beyond a mere spectrum purchase consideration. As described earlier, the Digital Dividend implementation will be heavily influenced by how regulators in each country choose to award the spectrum, and will force operators to evaluate a wide range of interlinked strategic decisions: Strategic positioning: operators will need to think about which customer segments they want to target, what the best operational model is to do so, and the positioning that will best support this strategy (e.g. no-frills price-oriented player vs. premium-based best network positioning, or somewhere in between). It is not self-evident that an operators current strategy will remain valid in the context of the new paradigm brought about by the Digital Dividend Commercial proposition: related to the above, operators will need to carefully assess the degree and scope of convergent value propositions, and the extent to which regional differences or a multi-brand strategy might be advisable. This is particularly relevant for integrated operators, which may see their fixed customer bases cannibalised by the new services enabled by the Digital Dividend Technology strategy and choices: while many may assume that the Digital Dividend and LTE need one another, it is not necessarily the case. In the face of potentially inflationary auctions and unclear vendor technology roadmaps, some operators may choose to deploy LTE (or even stick to HSPA+) on alternative bands to the 800 MHz and forego the Digital Dividend altogether, even if that results in a slight time to market disadvantage versus the competition. The issue of spectrum refarming is critical in this context Partnerships: we expect to see an increase in M&A activity, network sharing, domestic roaming agreements and other types of strategic partnerships between operators as a result of the Digital Dividend, as operators look for creative ways to ensure a positive return on investment for their shareholders. In some cases, these partnerships will be formalised prior to the spectrum auctions (e.g. one of Swedens Digital Dividend auction winners, Net4Mobility, is a 50/50 joint venture of Telenor and Tele2 created specifically with the goal of providing mobile broadband services in Sweden), and in others, they will come about depending on the auction outcomes in each country

Once again, the Digital Dividend appears as a trigger of a perfect storm implying a potentially big paradigm shift for telecom operators. The aforementioned strategic decisions and their relevance as moments of truth will force operators to evaluate, and eventually implement a Digital Dividend strategy in a cross-functional way. When it comes to the Digital Dividend, all departments within an operator (corporate strategy and development, marketing and sales, legal and regulatory, finance, and of course network and IT) will be affected by and contribute towards the implementation of a new broadband strategy.

11

EXHIBIT 8: TYPES OF DECISIONS AND DEPARTMENTS INVOLVED IN THE DIGITAL DIVIDEND

Source: Delta Partners

Given the vast financial, operational and strategic implications that decisions related to the Digital Dividend will bring along, operators would be well advised to tackle this issue early, well in advance of the spectrum auctions planned across Europe in the next 12-24 months (see exhibit below), to make sure their bidding strategy and network rollout plan are well informed and aligned with the business opportunity. As we will see next, there isnt a one-size-fits-all approach with regards to the Digital Dividend for all types of operators. The financial valuation of the opportunity is technically complex, which heightens the need for European operators to define their strategy as a matter of urgency.

EXHIBIT 9: LIKELY DIGITAL DIVIDEND EUROPEAN AUCTION ROADMAP

Source: Regulators, Delta Partners estimates

12

2. How much is the Digital Dividend spectrum worth?

Depending on the market position of an operator, the perceived value of the Digital Dividend may be very different, and thus the Digital Dividend can be seen as a threat or an opportunity. Factors such as the effect that the Digital Dividend will have on ARPU and churn given the new competitive landscape, the potential danger (or opportunity, depending on the operators point of view) of fixed line cannibalisation, and the mid to long term changes in market share position that the Digital Dividend may unleash, will heavily influence the value that an operator assigns to the spectrum. In order to illustrate the above, let us consider three generic operator types: Large integrated incumbents: these are operators with the largest market share and network in their countries, with ARPUs typically above average, and a full suite of services including fixed, mobile, broadband and often Pay TV. These operators typically have a large amount of GSM 900 spectrum Early entrant challengers: these are often large mobile operators occupying the number 2 or 3 spot in their markets, with a fair share of GSM 900 spectrum, and providing primarily mobile services (sometimes also fixed and Pay TV). These operators are sometimes squeezed between the incumbent and the late entrants and struggle to maintain market share Late entrant challengers: these are number 3-5 operators in their countries, with growing customer bases and market share, but a less favourable spectrum position than their competitors. In many cases, the GSM spectrum portfolio of these operators consists only of 1800 MHz, although they may also have 2100/2600/3500 MHz frequencies. In some extreme cases, these operators do not have any GSM spectrum at all and rely on domestic roaming agreements to provide GSM services to their customers. These operators are the most likely to compete on price

EXHIBIT 10: EUROPEAN MOBILE OPERATOR CATEGORISATION

Source: Delta Partners

13

Each of these operator types have to weigh different considerations when valuing the Digital Dividend spectrum. It is reasonable to assume that late entrants will find more upside in the Digital Dividend opportunity than a well-established incumbent, which might see its comfortable competitive position threatened. As a matter of fact, the value of the Digital Dividend spectrum for an incumbent operator may be purely strategic, and some incumbent operators may be content with acquiring this spectrum only to prevent their competitors from getting their hands on it.

EXHIBIT 11: IMPACT OF DIGITAL DIVIDEND ON KEY BUSINESS DRIVERS

Source: Delta Partners

3. Technology and strategy: a prisoners dilemma

Aside from the strategic implications that gaining access to new spectrum can have for operators as they define and implement their broadband strategies (todays telecom Nirvana), there are key network related decisions such as refarming of frequencies and the rollout of new technology bearers including LTE and/or HSPA+ - that, as a collateral effect, have the potential to drive strategic partnerships among operators, in their simplest form through network sharing, and in certain cases through full mergers. This effect may be amplified or accelerated depending on how regulators manage the Digital Dividend process, as we saw in the previous chapter. Delta Partners predicts that in most countries, three equally sized 2x10 MHz Digital Dividend licences will be awarded. There is little that operators can do with a smaller frequency block, and we generally expect regulators to impose a 2x10 MHz spectrum cap to ensure a minimum level of competition. In both Germany and Sweden, the Digital Dividend spectrum was split into six blocks of 2x5 MHz, and the end result of both auctions was that three bidders acquired blocks of 2x10 MHz each. This outcome is likely to repeat itself across Europe. Having the most to lose from letting competitors acquire digital Dividend spectrum, we expect that in almost every case, the national incumbents will acquire one of the licences, and the other two licences in a given country will be acquired by early or late entrant challengers. This means that in countries with more than three operators, at least one of them will be left out without any 800 MHz frequencies, unless they partner in advance and participate in the auction as a single entity. In the case of Germany, E-Plus failed to acquire Digital Dividend spectrum, and in the case of Sweden, Tele2 and Telenor participated jointly through their 50/50 joint venture Net4Mobility.

14

Under these circumstances, the three operators with Digital Dividend licences may need to come to terms with the notion of partnering with one of its peers to compete more effectively, although it is not self-evident that partnering for network sharing will always be value accretive. Building a shared network could provide two challengers a strong platform from which to erode an incumbents market share, compounding the benefit of Capex savings for the two partners. For incumbent operators, however, the Capex savings from building a shared network may in some instances be offset by a loss of market share, especially in favour of the weaker partner, and potentially also by a reduction in ARPU resulting from an eventual price war triggered by the remaining challenger or challengers, being forced to compete on price as the only way to fight bigger incumbents. Thus, it is crucial, especially for incumbent operators, to evaluate the pros and cons of network sharing from a strategic and financial perspective before starting partnerships discussions with other operators. In any event, in cases where the provision of mobile broadband services is clearly economically unviable, as might be the case in certain rural areas, operators may seek to share their networks on an ad-hoc basis in these regions.

EXHIBIT 12: ILLUSTRATIVE EFFECT OF NETWORK SHARING FOR INCUMBENT OPERATOR

Source: Delta Partners, based on actual project but sanitised to protect clients confidentiality

The second or third operator in a given market, typically the early entrant challenger in our earlier categorisation, may often hold the upper hand when it comes to partnerships, as it may benefit from partnering with either the incumbent (in order to marginalise smaller players), or one of the smaller players (to eat into the incumbents lead). This is a rare instance when being a middle child may actually be advantageous, and may prompt a prisoners dilemma situation for the incumbent, which may be better off in a market environment with no partnerships, but for whom the risk of two smaller operators partnering is larger than the economic downside of the incumbent partnering with another operator to pre-empt a coalition of the weak. In the example illustrated in exhibit 13, the highest combined pay-off for the incumbent and a challenger operator happens when neither partners with a third operator, but an alternative equilibrium point exists when both partner to avoid a potentially worse individual pay-off. Of course, the two operators can only partner with someone else if there are at least four operators in the market. In a market with three operators, the middle operator may end up being a much sought-out partner for both the incumbent and the smallest operator as each tries to prevent the other from striking a partnership.

15

EXHIBIT 13: IMPACT OF NETWORK SHARING PARTNERSHIP FOR INCUMBENT AND CHALLENGER OPERATOR

Source: Delta Partners

4. Alternatives to the Digital Dividend

The Digital Dividend spectrum is not the only avenue for operators to provide mobile broadband services to their customers. While lacking in terms of relative performance compared to LTE, HSPA+ is a proven technology, still evolving, and data throughput will continue to increase in the coming years (actual tests have been performed demonstrating HSPA+ data rates above 100 Mbps). A more intriguing possibility is the rollout of HSPA+ or LTE in the 900 MHz band. This would of course entail refarming the 900 MHz spectrum and reconfiguring the GSM infrastructure to cope with a hybrid GSM/HSPA+/LTE radio network. Among other things, this would include optimising and sectorising (if needed) the surviving GSM base stations to avoid interference with LTE, and adding new GSM base stations to make up for the capacity loss. In addition, in a refarming scenario, more LTE sites would be required than in a pure Digital Dividend scenario because of the limited 900 spectrum available for LTE and the need to reuse frequencies. However, an LTE 900 site would be on average 5%-10% cheaper than an LTE 800 site because of the possibility of reusing some of the passive infrastructure elements of the existing GSM 900 sites. All in all, spectrum refarming is likely to result in higher cumulative Capex for operators than rolling out an LTE network in the Digital Dividend band. It is therefore important for operators to consider two aspects when evaluating the spectrum refarming option: Is the incremental Capex lower than the likely licence fees for the Digital Dividend spectrum? Considering the time required to refarm the spectrum and procure LTE 900 compatible equipment, will competitors that focus on an LTE 800 network enjoy a material time-tomarket advantage?

16

EXHIBIT 14: CUMULATIVE 10-YEAR CAPEX FOR LTE 800 AND LTE 900 (REFARMING) SCENARIOS

Source: Delta Partners, based on actual example but sanitised to protect clients confidentiality

5. How operators should approach the Digital Dividend process

As we have seen, the Digital Dividend not only urges operators to make very relevant Capex and network related decisions, but also has the potential to disrupt the competitive dynamics in the market and ignite a new wave of consolidation and partnerships. As such, operators need to act cautiously by designing a strategic roadmap that optimises their own business case. From a large integrated incumbents point of view, the best possible scenario is the current status quo, favouring a divide and conquer strategy where the Digital Dividend auction is delayed as much as possible, the current GSM 900 licences are extended, and eventual Digital Dividend licence coverage requirements are either non-existent or technology neutral (since most likely the incumbents are already meeting these coverage requirements with their 3G networks). If push comes to shove, and assuming this is possible without infringing licence obligations, incumbent operators may opt to buy Digital Dividend spectrum but do nothing with it or just meet the minimum coverage requirements, in order to prevent the competition from acquiring this spectrum. Early entrant challengers, on the other hand, might pursue a regroup and jump strategy that allows them to leap forward and eventually (co)lead the market. To do that, they might need to partner with another operator to share the network and realise Capex savings, operational synergies and a faster time to market in order to more effectively compete with the incumbent (Oranges and T-Mobiles joint venture in the UK, Everything Everywhere, is a good example of such type of partnership).

17

Finally, late entrant challengers, especially those that do not own GSM 900 spectrum, might push for a big bang strategy that allows them to acquire 800 MHz spectrum at all costs and partner with other challengers. In addition, they also need to lobby the regulator to re-auction the GSM 900 frequencies sooner rather than later, in order to improve their spectrum position at the expense of the incumbent and early entrant challengers. This may be a late entrant challengers last chance to become a relevant market player. All in all, European operators need to understand the competitive scenarios that may develop in their markets as a result of the Digital Dividend, some of which they can influence, while others they may be forced upon. Next, operators should value each scenario and then define and execute a lobbying campaign to promote the scenario that best meets their strategic objectives. Finally, and ideally when all the auction design elements are known, operators should perform a battery of war game and auction simulations to inform their bidding and competitive strategy.

The operators moment of truth to unleash or avoid The Perfect Storm

We expect the European Digital Dividend process to heat up after the first two auctions in Germany and Sweden have been completed. As we saw earlier, most of the Western and a few of the Central European countries will auction the Digital Dividend spectrum in the next 12-18 months. The remaining European countries (and the rest of the developing world) will follow from 2013 onwards, as the analogue to digital TV switchover is slated to take place later in those countries. Given the complexity of the decisions involving the Digital Dividend, the disparate interests of the constituents involved, the financial impact for all players and the Digital Dividends role as a catalyst of a new competitive landscape, operators in the countries where spectrum auctions are scheduled in the next 18 months need to define their strategic game plan now in order to capitalise on the opportunity, assess the key technological options in response to or as a driver of a new competitive landscape, and strike appropriate partnerships early, anticipating their competitors moves. Billions of dollars in licensing fees and Capex as well as in additional revenues are at stake. Winners in the Digital Dividend process will come out stronger than they are today, while losers may be forced out of the market or be severely handicapped.

18

A call to action for operators

The complexity and breadth of decisions that CEOs of telecom operators will need to make in the wake of the Digital Dividend require a well coordinated, modular and phased approach to ensure success.

Just like captains of boats and airplanes study their navigation charts and weather forecasts ahead of their journey to avoid or protect themselves against inclement weather, so do telecom CEOs must prepare to deal with different kind of storms. This is especially the case in Europe today given the magnitude of the storm that the Digital Dividend has the potential to unleash.

19

Delta Partners is the leading TMT advisory and investment firm in emerging markets. With more than 160 professionals, the firm operates across 50 markets in the Middle East, Africa, Central & Eastern Europe and Emerging Asia. Delta Partners provides three synergistic services: management advisory, corporate finance and investments from its offices in the UAE, Bahrain, South Africa, Spain and Singapore.

Advisory: Delta Partners advisory professionals partner with C-Level executives in telecom operators, vendors and other TMT players to help them address their most challenging strategic issues in a fast-growing and liberalising market environment in over 50 markets.

Investments: As a fund manager, Delta Partners manages an $80Mn private equity fund, targeting investment opportunities in the TMT space in high growth markets. The focus is the Middle East, Africa, Eastern Europe and Emerging Asia. Delta Partners private equity fund leverages the firms unique TMT industry expertise to create value for its investors throughout each stage of the investment cycle, from deal sourcing to supporting portfolio companies in driving value extraction.

Corporate Finance: Delta Partners provides corporate finance services and has been involved in several buy-side and sell-side telecom transactions in the region. As true industry specialists, the firm offers a differentiated value proposition to investors and industry players in the region. Delta Partners actively leverages its close link to its private equity arm to access the investor community as well as top-level financial talent.

Delta Partners delivers tangible results to its clients and investors through its exclusive sector focus on telecom, media and technology, and a unique approach to services, combining strategic advice and a hands-on pragmatic approach.

For more information about Delta Partners please visit: www.deltapartnersgroup.com For a list of all Delta Partners white papers please visit: http://www.deltapartnersgroup.com/our_insights/whitepapers

Copyright 2011 Delta Partners FZ-LLC. All rights reserved.

20