7. Recommended methodology

Introduction

7.1 The preceding chapters have outlined how the problem of determining the market value of the rights might be considered. The methodologies have been considered on a generic basis.

7.2 A key issue for this exercise, however, is the level of confidence that a calculated value will indeed approximate the market value, as achieved at auction if the right holders choose not to exercise their renewal rights. As previously noted, the ability of any particular methodology to proxy market value will depend on a large range of factors, many of which are likely to be subjective. The availability and robustness of information and the time and resource available to undertake the exercise are particularly important.

7.3 The obvious approach to assessing this question would be to run a simulated auction. However, this is likely to present some significant challenges, in that inputs to the simulated auction would need to be developed in relation to the:

• potential bidders;
• maximum and minimum valuations of each bidder, which will depend on their business strategy and plans;
• way in which maximum and minimum valuations would change, depending on other bidders and their behaviour in the auction;
• probabilities of entry to the auction;
• spectrum packages offered in the auction; and
• alternative means of entering the market.

7.4 As noted in Chapter 4, we judge the most probable outcome as being that there would be no new entrants. However, this cannot be said with certainty and the emergence of a potential new entrant on the cellular market could encourage strong bidding by incumbents to prevent entry and lead to potentially high prices for the spectrum. Likewise, we cannot predict with certainty whether there would be competition between the two incumbents (and the extent of that competition). Press reports at the time of writing have speculated that, following Telstra’s decision to move away from the CDMA family in Australia, Telecom might need to build a new cellular network using the GSM family. In this case, it might be in Telecom’s interest to acquire some spectrum in the 900 MHz band.

7.5 Given uncertainty about whether there would be new entry or not, it would be necessary to develop the inputs to the simulated auction noted above. This would require use of one or more of the valuation methodologies noted in the previous chapters. It would also be an increasingly subjective and difficult exercise, typically requiring vital information about the likely attitudes and behaviour of the incumbents.

7.6 We discussed this approach with the Ministry. The Ministry advised that the complexity of the exercise and the time required for its implementation were neither consistent with its objectives of a simple licensing process nor with the timetable for the renewals process.

7.7 We suggest that a pragmatic approach to dealing with this issue is to select a single methodology, to form the core of the value calculation. Judgemental adjustments may be required, but without necessarily going to the lengths of running a simulated auction.

7.8 The conclusions from the preceding chapters are that an incremental ODV methodology, applied at the margin, would logically be used to determine a price in a competitive market, with a restricted supply of spectrum.

7.9 The choice of incremental ODV as the core methodology reflects a need to be pragmatic about the trade-offs between accuracy of methodology and practicality of implementation. The incremental ODV methodology will require a relatively limited set of relatively readily available information, compared to that required to implement, for instance, a simulated auction.

Outline methodology

7.10 We propose the following methodology to determine the price to be offered to the existing right holders.

Incremental ODV methodology

7.11 Estimation of the costs avoided (i.e. the cost required to reconfigure the network to maintain a given level of service following a marginal reduction in spectrum holding) can be undertaken either by assessing the actual technical requirements and associated costs, or by modelling.

7.12 Assessment of the technical requirements (and costs) involves examining the specific network configurations of the existing spectrum holders, which are likely to differ significantly from each other. However, the aim of the pricing exercise is to determine a spectrum price that relates to a “generic” operator, pointing to modelling as the most appropriate valuation method.

7.13 The basic approach can therefore be summarised as:

• Specify scenarios for expected levels of network traffic, by year.
  • Break the country down into bands of traffic density (and hence network/spectrum requirements).
• Specify service quality levels (call blocking rate).
• Determine the optimal, modern-equivalent network technology (this also specifies the minimum spectrum package).
• Determine the optimal network configuration.
  • “Optimal” is defined for these purposes as “lowest cost.”
• Determine the spectrum package required to support the optimal network configuration.
• Scope the deprival scenario.
  • Specify the incremental spectrum package of which the operator will be deprived.
  • Specify the optimal (lowest cost) network configuration for the reduced spectrum holding.
• Determine the difference between the two scenarios.
  • Change in number and/or type of sites required and any other changes in assets.
  • Change in operating and maintenance activity.
• Apply average or standard costs to each quantified change, to determine the change in costs, and hence the marginal cost of the unit of spectrum.
• Apply probabilities against the scenarios and perform an expected value calculation on these, discounting the change in costs to the time of renewal, at a generic operator’s cost of capital.

7.14 This will produce a figure per unit of spectrum. This figure will be applied to all rights.

7.15 It should be remembered that valuation is an art, rather than a science, and that there will be a significant element of judgement to all the above calculations and steps in the process. Unfortunately, in the final analysis, all that will be certain is that if the operators accept the offer price, then it is lower than they were actually prepared to pay.

Benchmarking

7.16 A valuation exercise would normally seek to use more than one single methodology. Therefore, although we have dismissed benchmarking as a primary valuation approach, we recommend that reference should nonetheless be made to the available New Zealand and international transactional data. The aim of such an exercise would be to assess whether the figure calculated from the ODV methodology lies within a reasonable range.

7.17 Benchmarking needs to take account of:

• Differences in the relevant markets
  • Market size
  • Level of competition
  • Relative pricing
  • Relative wealth
  • Maturity
• Geographical differences:
  • Population density
  • Terrain
• Exchange rates
• The age of the benchmark (i.e. how recent it is)

7.18 The benchmarking data could be used to develop a range of spectrum values to provide external points of reference for the ODV calculations. These could also be evaluated for meaningful trends in spectrum values. The use of the benchmarks will involve a significant element of judgement.

Further adjustments

7.19 This assignment has not involved actually performing an ODV calculation or benchmarking exercise. When these exercises are undertaken and the results become available, consideration will need to be given to the extent to which the calculated value is likely to approximate in practice the results of an auction of the spectrum.

7.20 The ODV methodology becomes more robustly applicable with increasing numbers of buyers and sellers in the market (as noted earlier in this paper). When considering the actual renewal price to offer, it will be necessary to exercise judgement as to whether the calculated figure needs to be adjusted to reflect the particular characteristics of the New Zealand market.

Variants

7.21 Of the four items noted above, the ODV exercise, with a short time horizon, is primarily an engineering exercise: determining the optimal network configuration to serve a given quantity of demand at a given service level.

7.22 If considered over the longer term, issues of growth in cellular traffic and changes in pricing become more pertinent and the forecasting of these becomes less certain as the time horizon is extended.

7.23 This issue could be addressed by either:

• Recalculating the ODV methodology periodically (e.g. the UK resets its Administered Incentive Pricing every three years); or
• Renewing the rights for a shorter period of time.

7.24 If the ODV methodology is applied and pricing re-applied every five years, it is likely that major technological or market changes will be reflected in the updated pricing. This would eliminate the need to consider technological uncertainty up front.

7.25 Premia or discounts do however need to be calculated for the additional value in having certainty that new entrants will not be able to enter the market and to reflect potential uncertainties of an auction. These are much more subjective factors than the ODV methodology. It would make sense that they are applied only once, at the point of renewal.

7.26 Any reduction of the renewal period would also need to be carefully considered, lest it undermine the basic policy intent: the Ministry is seeking to minimise economic disruption by providing certainty of renewal to the existing right holders, should they wish to do so.