The draft National Telecom Policy (NTP) 2011, released by the Ministry of Communications and IT in October 2011, gives a major thrust to transparency, improvement of the investment climate and promotion of consumer interest. The policy is likely to bring some relief to the sector, which has been caught in a series of controversies regarding spectrum and licence issues, besides facing hyper-competition and declining revenues. In this context, tele.net organised a conference, “Impact of National Telecom Policy 2011”, on November 15, 2011. The conference covered key issues such as spectrum pricing and allocation; one nation, one licence; consolidation; and broadband. The following section presents the highlights of presentations made at the conference...

Kunal Bajaj, Director, Analysys Mason India

Several policy and regulatory initiatives proposed in the NTP 2011 are likely to have long-term implications for the telecom industry. A look at the possible impact of the proposed changes in the spectrum limit, pricing and allocation...

Spectrum limit

The maximum spectrum holding limit for a post-merger entity is recommended to be increased from the current 15 per cent to 25 per cent of the spectrum assigned in the particular circle. This is a positive move as it will allow operators who currently hold about 15 per cent of the spectrum in a circle (regardless of how much spectrum is available per circle) to acquire or merge with another player. Limiting the spectrum at a certain percentage level enables the total amount of airwaves held by an operator to be raised with additional spectrum being allocated in that circle.

In Mumbai circle, for instance, a total of 107.4 MHz of frequency division duplex spectrum has been allocated for telecom services. The proposed spectrum holding limit implies that an operator will be able to hold about 26.5 MHz of spectrum in the circle. At present, the highest amount of spectrum held by any operator in the country is 20 MHz (2G and 3G), which is held by Mahanagar Telephone Nigam Limited (MTNL) for the Delhi circle.

In many developed and emerging countries, operators have more than 25 MHz of spectrum, while their subscriber base is less than that of India’s leading operators. Higher spectrum allocation helps operators to reduce their opex and capex for offering wireless services as they do not require additional base transceiver stations (BTSs) to enhance network capacity.

The recommendation to allow operators to hold up to 25 per cent of the total spectrum allocated will facilitate industry consolidation. Also, the 25 per cent mark ensures that there will be at least four operators in a particular service area and discourage spectrum hoarding/blocking.

Spectrum pricing

Compared to the pricing in 2008, spectrum charges (for up to 6.2 MHz) are proposed to be increased by 6.6 times from Rs 2.91 billion to Rs 19.24 billion per MHz. For spectrum in excess of 6.2 MHz, it has been proposed to increase charges by 17.1 times from Rs 2.91 billion to Rs 49.7 billion.

Proposed one-time charges for spectrum beyond 6.2 MHz will have negative financial implications on incumbent operators. Based on the current spectrum allocation, Bharat Sanchar Nigam Limited (BSNL) and MTNL will have to pay a high premium in the form of one-time excess spectrum charges. Among the private players, the highest spectrum charges will be about Rs 38.2 billion for Bharti Airtel, Rs 22.85 billion for Vodafone, Rs 14.45 billion for Idea Cellular, Rs 11.5 billion for Aircel and Rs 0.85 billion for Reliance Communications (RCOM). Therefore, for several players the one-time excess spectrum charges are almost equivalent to the 3G spectrum prices.

Moreover, there are several issues in the methodology that the Department of Telecommunications (DoT) needs to consider before accepting the Telecom Authority of India’s (TRAI) recommendations. The cash flow and substitution models adopted for calculating the current value of spectrum do not account for quality of service or service innovation. The Cobb-Douglas function used to calculate the prices has several drawbacks, for example, it ignores revenues (assumes that demand is inelastic) and some important cost elements (such as licence and spectrum fees).

The prices are not benchmarked against spectrum auction prices in other countries. While this can be justified by the uniqueness of the Indian market, the calculated price of $0.85 per MHz per population in India is high as compared to global standards.

There are certain key assumptions around network capex, ARPUs and BTSs in the model, and any modifications in these can significantly impact the final value. For example, the model assumes a lifetime of 20 years for network assets, which should be less than 10 years.

Spectrum refarming

TRAI’s recommendations on 800/900 MHz spectrum refarming may have a negative impact on the incumbent operators who hold low frequency spectrum in the majority of circles. The earning before interest, tax, depreciation and amortisation for an operator (offering services for 10 years) in the 1800 MHz band is up to 100 basis points lower than that for an operator in 900 MHz band. This is because of the larger number of towers required (up to 1.5x) to provide adequate coverage.

Several countries have implemented spectrum refarming, but the approach has been different. Reassigning 2G spectrum upon licence expiry is the most popular approach adopted, especially in European countries. It includes both upgrading the technology to be used on the spectrum with the same operator as well as reassigning only a part of the spectrum. The measure to take away the entire 900 MHz spectrum has not been adopted in any country so far, especially where the existing operators did not have a right of first refusal. Most of the countries have followed an auction process to reassign spectrum.

Spectrum sharing, pooling and trading

The draft policy envisages greater infrastructure sharing, including pooling, trading and sharing of spectrum. This will enable operators to reduce network costs and improve profitability.

Although the operational details and timelines of the initiative have not been mentioned, spectrum sharing can generate substantial cost savings for the industry as a whole. Operators can generate up to 52 per cent savings in capex on a per erlang (Er) basis in a two-operator-sharing scenario. (Assumptions: capex per Er of $700, capacity per site of 34 Er in a single-operator scenario and spectral efficiency of 2.1x resulting from spectrum sharing in a two-operator scenario.)

However, spectrum trading presents the risk of anti-competitive activities such as spectrum hoarding/blocking and speculative purchase of spectrum from non-operating players. With spectrum sharing, operators will no longer be able to differentiate themselves on the basis of quality of service. Technological innovation may also be impacted as changes in the radio access network require more coordination.

Spectrum road map

The overall concept of having a road map is important as the carriers need to know what the network rollout scenario will be in the future. A spectrum road map every five years will give industry players greater flexibility for network planning. The release of additional 500 MHz by 2020 will help operators augment network capacity.

However, uncertainties regarding timelines and the operational process still exist. Certain areas, such as white space and digital switchover, have been missed out. A clear road map for refarming spectrum from non-commercial entities also needs to be developed as its success and future spectrum allocation will depend on the effectiveness of such refarming.