6. Assessment of valuation approaches

6.1 Set out below is an assessment of the valuation approaches relative to the Ministry’s objectives. This is followed by a discussion on the approaches. More detail on the evaluation is contained in Appendix C. Particular attention is paid below to the Covec approach and its applicability in the context of the cellular telecommunications market, as opposed to, say, the broadcast markets.

6.2 The criteria used in the evaluation accord with the Ministry’s objectives for the development of a pricing methodology. In particular, whether the approach:

• Is transparent
• Will be relatively simple to apply in practice
• Will produce an outcome that has a reasonable probability of being a reliable proxy for market value. Table 6 Summary Evaluation of Valuation Approaches

Table 6 summary evaluation of valuation approaches

6.3 The approach to assessing whether each methodology can approximate market value has been to focus principally on whether, in the case of the cellular market to which this exercise relates, there is likely to be a notable flaw in the underlying methodology. We have not attempted to rank each methodology in terms of its potential accuracy in estimating market value. Such a ranking would in practice be greatly influenced by the availability and robustness of information and the time and resource available to undertake the exercise. Such factors are reflected in the transparency and simplicity scores.

 Top

Benchmarking

6.4 Ideally, in determining the market value for the spectrum rights, recent, relevant and observable benchmarks would be used. The more recent and relevant, the closer the benchmark might be assumed to be to approximating current market value. We noted in Section 2 that ideally, market value would be determined by reference to observations of the price at which the asset being valued or similar assets are traded in an active market.

6.5 We do not consider that benchmarking is appropriate as the primary valuation approach. The available New Zealand benchmarks range from five to fifteen years old. The market for cellular services has changed dramatically over that time. With the current introduction of new data-intensive services, the market can be expected to continue to change in the future.

6.6 Overseas benchmarks could be used. However, they are likely to suffer problems of translation to the New Zealand context. They might also be difficult to observe – putative values for spectrum rights might not be available from trades in spectrum, but need to be derived from financial data available from, for instance, merger and acquisition stock market and regulatory filings.

Earnings valuation

6.7 The earnings “valuation” methodologies are not all ideal either:

• A full business valuation will produce a robust estimate of market value but is likely to be complicated. It will require an apportionment of the business’ value across its assets (tangible and intangible), introducing further subjectivity.
• The simplified approach could be a workable solution but like the full valuation approach, it will suffer from the need to apportion value across all assets, including intangible assets, to derive the value for spectrum rights.
• Covec suffers from the same issues as benchmarking (or use benchmarks of doubtful relevance as its starting point and hence produce doubtful outcomes).

6.8 The Covec approach is the methodology applied by the Ministry in the renewals of rights for radio and television broadcasting rights. The general approach taken by Covec to the price setting formula is to:

• Use the original price paid at auction for a spectrum right as a base price; and
• Apply to the base price valuation a simple growth factor (referred to by Covec as "z") which is assumed to reflect the change in value of the spectrum between the original auction and the reallocation.

6.9 This approach has the advantage of simplicity, but does rely on:

• Being able to identify and calculate a growth factor that is a suitable proxy for the change in value – typically this is taken in terms of growth in net cash flows (from which the value of the spectrum right is ultimately derived);
• Assumptions that the:
  • growth factor is essentially the same for each right holder to which it is applied (i.e. market shares are constant over time);
  • growth factor is stable over time (i.e. there has been and will be constant growth over the 40-year period⁶ under consideration); and
  • original auction prices were "correct" (i.e. that the expectations on which the auction bid was made have proven to be accurate as the market has developed over time).

6.10 The Covec formula uses changes in industry revenues over a 40-year period as the growth factor. For changes in revenues to be a suitable proxy for changes in net cash flows it must be assumed that profit margins are constant over the relevant period.

6.11 Given the assumptions outlined above, the approach is therefore best suited to situations where:

• Robust data on historical industry revenues is available;
• It is possible to forecast future revenues with some degree of certainty;
• The industry is relatively mature and not subject to dramatic change from technology or some other factor and therefore:
  • revenues can be expected to follow a relatively stable growth path over the long term; and
  • the original auction prices reflected expectations that have proven to be reasonably correct (i.e. the base prices are still valid in relation to the current and future market).

6.12 Being realistic, the tests outlined above are hard to meet for any industry. However, there is a higher level of confidence in their applicability in the broadcast market than in the cellular telecommunications market:

• There is limited data publicly available on historical cellular revenues for the 1990s.
• The cellular industry is now more mature (at least in terms of simple voice telecommunications services), but was very immature when rights were initially auctioned, unlike the television and radio broadcast industries. There have been significant changes in the market over the last 10 years, both in terms of market penetration, prices and products. Anecdotally, we understand that the expectations of the cellular operators at the start of the market were very different to what has actually transpired. Therefore, we have a very low level of confidence that the original auction prices are a robust starting point for the application of the Covec formula (i.e. that the expectations on which they were based have not been borne out).
• The cellular industry also faces significant future uncertainty, with the introduction of high data-rate services (3G). Voice services are likely to remain an integral part of many mobile service offerings, but mixed with service or pricing packages that are much broader in scope. Forecasting future revenues is therefore challenging. Forecasting would involve assessing service volumes and prices. It would also be necessary to distinguish between cellular services that are spread across different parts of the spectrum. (This problem is not unique to Covec – all assessed valuation approaches face this uncertainty.) In relation to Covec, this and the prior point mean that there can be a much lower level of certainty in identifying long-term proxies for growth in net cash flow.

 Top

The deprival approach

6.13 The deprival approach is attractive. Its calculation can be systemised and made relatively transparent. The technical implications of deprival of spectrum should be an engineering determination that, once agreed, can be transparently costed.

6.14 In addition, deprival value is founded in economic principles. In a competitive market, without “lumpiness⁷” the marginal price (and the price at which the market will clear) of a good or service will equal the marginal cost of provision of that good or service. This suggests that the market price or market value of a unit of spectrum should be equivalent to the costs avoided by not having the marginal spectrum package. The maximum that a network operator would pay for the marginal unit of spectrum is the greater of the costs that it would avoid, or the net gain it would obtain by having the additional package. Similarly, the minimum at which it would sell the marginal unit of spectrum is the costs that it would incur, or the gain it would lose by foregoing that unit.

6.15 Consistent with the market definition discussion in Section 2, deprival value is based on a set of implicit assumptions about the competitiveness and depth of the market. Specific assumptions would include:

• Spectrum rights are freely tradable and usable by cellular operators in small units.
• There are no barriers to entry or exit to the market.
• The market has many participants, who are willing, non-anxious buyers and sellers.
• No participant exerts market power.
• There are no information asymmetries.
• Above a minimum amount of spectrum, network configuration and spectrum are generally substitutes (i.e. an operator is able to build a network to deliver a given level of service with a single spectrum package, or with many) and hence the amount of spectrum, above the minimum, does not affect its revenues (although there may be a significant impact on costs);
• The supply of spectrum rights is finite.
• Cellular services remain the highest-value use of the spectrum in question (i.e. cellular operators are competing with each other for spectrum, but not also with e.g. television broadcasters).
• Buyers and sellers in the market are profit maximisers. They use and trade spectrum rights to maximise their own profits.

6.16 Clearly, a number of the above do not necessarily apply, either in the mobile market globally, but more importantly in New Zealand. In particular:

• spectrum does not appear to be scarce in New Zealand, at least in the medium to long term;
• with only a single seller of spectrum and two network operators, it would be hard to characterise the market as “deep”;
• the level of competition in the cellular market can be expected to be lower than in markets with more operators.

6.17 As indicated in Section 2, the competitive market valuation needs to be adapted to reflect the specific conditions faced in New Zealand. In particular, as noted in previous chapters, if offered in an open sale (presumably by auction), there is a potential that no other bidders would emerge, and therefore the rights would be sold at the (immaterial) reserve price. Alternatively, the incumbents might be prepared to pay to protect their existing market position.

6.18 The deprival approach would involve forecasting traffic volumes that a network needs to accommodate. As noted above, this is not a simple task.

 Top

Right holders’ views

6.19 We have held a number of discussions with Telecom and Vodafone and also had the content of some discussions between the Ministry and the right holders relayed to us.

6.20 Where opinions were expressed, right holders were generally in agreement with the assessment of the potential valuation methods noted above in Table 6.

• Benchmarking was seen as difficult, particularly the translation from the overseas population, population density, geographical, regulatory and competition contexts to New Zealand. The Nordic countries were offered as the most comparable to New Zealand in terms of population densities and variability and ruggedness of terrain.
• Valuations were considered to be highly subjective, particularly given the potential timescale that financial models would need to consider.
• Both right holders were opposed to the concept of a full deprival value, as well as considering it to be highly complex. One right holder proposed the exploration of Administered Incentive Pricing as a potential approach to the pricing problem.

6.21 Two key themes also emerged clearly from the discussions.

• Simplicity is important: there is a preference for a simple approach; and
• Certainty as to renewal pricing and the consequences of not accepting the renewal offer (in terms of the timeframes and methodology for the auction) is highly valued.

6.22 One right holder also expressed the view that whatever the renewal or auction methodology, it was necessary to ensure that there was a financial commitment involved at the point of renewal or auction. The holder should not be allowed to obtain rights to the spectrum without a commitment.

⁷Here, “lumpiness” means any technical or economic impediment to a continuous pricing spectrum. In this case the need to have at least a minimum degree of coverage before the spectrum can be regarded as valuable in its own right is a form of lumpiness.