Thursday, December 5, 2013

For a Telecom Revival

Positive steps on telecom and broadband need to be bolstered with more for a resurgence

Shyam Ponappa   |  December 5, 2013

The government announced momentous decisions subject to Cabinet approval on telecom policy on December 3*. There are some major pluses: increased spectrum made available, and higher market shares allowed through acquisitions. Less constructive for the sector are decisions like acquirers having to pay for spectrum above a floor (4.4 MHz for GSM and 2.5 MHz for CDMA) at market rates unless the spectrum was won through auctions. While there are positive decisions, more are needed for true resurgence in this sector.

Perhaps there's also a need to curb the inappropriate application of direct-democracy to complex issues. This refers to choices influenced by uninformed but vociferous public opinion, whereas the requirement is for logical conclusions based on knowledge and understanding of the facts, domain expertise, and skill in problem formulation, solution design and implementation. The underlying constructs may include factors like technology; economics and its dissimilar sibling, finance; society's organisation, capacity and inclinations; and the law. This is especially true for infrastructure, a recognised weakness in our economy. The issue is that misdirected policies can result from the indiscriminate application of old frames of reference, customary practices, or just following the herd.

Consider the state of telecom and broadband: how bad our services are, and how badly the sector is doing, despite the enormous potential. Decisions on spectrum have profound effects on how these services affect productivity and living standards, with inappropriate policies resulting in impediments and misdirection. This is especially important in developing economies because the opportunity losses are unaffordable, and recovery is difficult in the absence of robust institutions and processes. Negative examples like the drive to refarm 900 MHz spectrum and maximising short-term government revenues from spectrum make a mockery of government-for-the-people. "Refarming" refers to mobile operators having to give up most of their 900 MHz band holdings for redeployment of newer technology, primarily because more developed economies did so. Existing operators would lose much of this spectrum, unless they win it back through auctions or acquisitions. This is like taking away captive mines from established steel manufacturers to create a "level playing field".

There are differences, of course, between spectrum and mineral resources. Unlike minerals, spectrum is not depleted by usage, the time taken to develop a new mine is usually more than to deploy a new network, and so on. But refarming will entail significant costs for new networks with many more base stations. This will take years, requiring interim arrangements to avoid service disruption to existing users. It seems like an enormous burden, in effect cross-subsidising newer technology for the high-end user segment. 

How bad is the situation for the industry? Take indicators like profitability, debt, and spectrum costs. Chart 1 shows Earnings Before Interest, Tax, Depreciation and Amortisation (EBITDA) as a percentage of revenue for mobile network operators in India, Indonesia, Malaysia, China, Thailand and Singapore, with India being lowest at 20 per cent.

Chart 1: Mobile Network Operators’ Profitability - APAC

Source - GSMA + BCG:

Regarding indebtedness, two Indian operators have Debt/EBITDA ratios at 4 and 6, well above acceptable levels. Others, whose debt is in line with Asian operators, are less able to service it because of lower revenues. While some urge that leveraged companies in difficulties be allowed to fail, the magnitude is such that there is a serious risk of destabilising the economy.

Spectrum reserve prices in India are much higher than in other countries (Chart 2), despite the average revenue per user (ARPU) being much lower in terms of purchasing-power parity (PPP), rendering investments unattractive.

                                            Chart 2

It is because of this stressed situation that the authorities, the industry, and the public need to reconsider their basic approach to spectrum needs. One reason for the forced refarming is supposedly that 900 mHz spectrum is needed for more efficient technologies. Another is that some operators with no 900 mHz spectrum are at a genuine disadvantage in terms of in-building coverage. Of course, the most compelling reason may be simply the government's need for revenues to cover its deficit, despite the enormous negative consequences to the long-term public interest. The question is whether there have been adequate efforts to explore less disruptive alternatives to achieve the objectives of reliable, inexpensive communication services.

Spectrum Bands & Ecosystems

As of May 2013, the prevalent frequencies in LTE networks in Asia were as shown in Chart 3.

Chart 3: Spectrum Bands in LTE Networks (Asia) May 2013

Source: Wireless Intelligence

The most common were 1800 MHz and 2.6 GHz networks. The 2.3 GHz band used in India (and China) is not very widespread, while 900 MHz is barely there. Bands that are not widely used are unlikely to benefit from scale economies. From this perspective, it is more logical to refarm 1800 mHz for LTE rather than 900 MHz, and the now widely adopted 700 MHz band.

The adoption of the APT700 band across Asia (including India), Latin America and Europe opens up the possibility of evolving into the largest LTE ecosystem with significant scale economies. As Verizon's established 700 MHz band in America differs from the APT700 band, the availability of devices may be a concern.

However, the fact that many countries have adopted the APT700 band improves the chances of quick development of equipment, starting with Telstra's planned trials in December 2013/January 2014.

For the resurgence of telecom and widespread access to broadband, the current positive moves to cut reserve prices somewhat, allow spectrum trading and consider uniform spectrum usage charges are not enough. Public opinion tends to view these steps as favouring telecom operators, or as sops to one operator or group. However, policies need to be formulated from considerations of the public interest, including that of users, the industry, and the government. Regarding auctions, there needs to be rethinking on the lines of the Swedish approach of bids for network investment and rollout, perhaps with incentives for faster delivery.

shyam [no space] ponappa at gmail dot com