0 Executive summary

The Ministry of Economic Development (MED) has requested a short study that examines:

• the optimal quantity of 800MHz and 900MHz spectrum for operators in New Zealand over the period 2011–2031:
  • the optimal quantity of spectrum required by the current management right holders in the 800 MHz and 900 MHz bands
  • the optimal quantity of 800 or 900 MHz spectrum that a new entrant would require in the 800 MHz and 900 MHz bands to operate a nationwide cellular network.
• the amount of 800/900MHz spectrum substantially used by the current management right holders.

Optimal spectrum in period 2011-2031

To determine the optimal spectrum we developed a bottom-up model that estimated the effect of splitting each of the 800MHz and 900MHz bands into two parts: one to remain with the incumbent (Telecom or Vodafone) and the other to be allocated to a new nationwide operator (TelstraClear or Econet Wireless). The model examined the amount of spectrum that should be allocated to each operator, varying in 5MHz steps between 0MHz and 20MHz. Defining the optimal split of spectrum as that which gives the lowest overall cost, we have determined that the optimal split is anywhere from a 5MHz:15MHz split to a 15MHz:5MHz split for both spectrum bands, assuming all operators are able to use their spectrum in higher bands. Telecom is unable to its higher band spectrum with its current CDMA technology; if this remains the case over the 2011–2031 period then the optimal split is 15MHz (Telecom):5MHz (TelstraClear).

The results show that it is very expensive to build a nationwide network without at least 5MHz of spectrum in the 800MHz or 900MHz bands.

Current utilisation of spectrum in the 800MHz and 900MHz bands

Using information provided by the operators to the MED and other publicly available data, we have found that the Telecom substantially uses 5MHz–8MHz in the AMPS-A band and at least 1MHz in the AMPS-B band. Vodafone appears to be substantially using all of its 900MHz spectrum.

We also undertook a benchmarking exercise as a means of comparing current spectrum amounts of all the operators in sample countries. This found that Telecom and Vodafone held more spectrum than nearly all the other operators within the sample. We considered whether there was some justification to warrant a relatively large amount of spectrum, due to the characteristics of the New Zealand environment. A regression model was used to ‘normalise’ the disparate operators within the sample for key factors that may influence the amount of spectrum awarded to operators. Applying New Zealand data to the regression model results in a estimated allocation of 2 × 9.4MHz for Telecom and 2 × 10.7MHz for Vodafone, which is less than the actual holdings of the operators.