Are telcos paying too much for 3G spectrum?

Since telcos have bid way above their comfort band, they need to get a much higher share of subscribers to use 3G services, and with significantly higher monthly billings - right now, both are a bit of a stretch.

Kunal Bajaj
Partner & Director India,
Analysys Mason Limited

With bids for 3G licences crossing five times the reserve price for some circles, the question is whether telcos are paying too much?

To make sense of the bids, you need to visualise how India's telecom market will look in a few years. By then, consolidation will have begun and only four or five large telcos will be left; 3G services will have been rolled out, 2G subscriber net additions will be in the low single-digit millions every month, and, most important, price will no longer be the prime differentiator.

In such a scenario, 3G will be a hygiene factor for surviving telcos, essential to retain any significant market share.

The valuation metric used globally for the evaluation of 3G auctions is dollar per MHz of spectrum per population covered. This metric reached as high as $4/MHz/pop for auctions in the UK, which left the participating carriers nearly bankrupt.

This metric has already crossed $4 for some markets in India (although at an all-India level, the numbers are a more sober $0.4/MHz/pop), and may result in a difficult business case for winners in those areas.

The value of spectrum is derived from three components. One, the ability to upgrade customers to 3G and derive higher average revenue per user (Arpu). Two, the revenue from subscribers who can be captured from competitors who do not win 3G.

Finally, the upside of retaining high-end customers who may have otherwise churned. In addition, the value of the spectrum is also augmented by the strategic value it holds owing to the scarcity of spectrum in India and uncertainty about when the next set of blocks will go under the hammer, and the actual market value of the spectrum asset itself.

While considering the value of retention, keep in mind that Indian telcos generate about 30 per cent of their revenues, and 45 per cent of Ebitda margins from the top 10 per cent of their user base.

Given that mobile number portability will be introduced later this year, losing telcos will be forced to concentrate only on the more basic users of 2G as they will slowly lose the majority of their high-end user base to new 3G licence holders offering bundled 3G smartphones, lots of free voice minutes and compelling video content.

In terms of attracting new subscribers, owning 3G spectrum becomes all the more important since there are a limited number of variables that can be used to differentiate a telco's offering, apart from price.

With vendor financing easily available, network-based differentiation is only mildly sustainable; customer service continues to be outsourced by a majority of telcos to the same types of call centres, and there is little else that customers really care about.

Using 3G as a platform to launch sticky services, like on-demand video, is what will allow telcos to add another dimension to their competitive arsenal.

3G will also lead to a revenue uplift from existing users. 3G networks are allowing increased flexibility for data-voice mix and deployment scenarios. Carriers can focus on a data-centric deployment in business districts and metro markets to drive sales of high-speed packet access modems (dongles) with Arpu of about Rs 800, and enhance the contribution of non-voice in their revenue mix.

Since 3G also allows for a tighter device service integration, a better user experience where applications and services are easier to find and use will also help increase Arpu, or at least prevent its further decline.

While this remains true for the existing high-end customers, even for the large middle-class customers, 3G connectivity through a mobile handset may be the primary means through which they can access high-speed internet for the first time.

Therefore, carriers also have the opportunity to develop an entire 3G-based ecosystem for driving mobile-internet usage and its adoption across urban and rural areas.

In the short term, 3G business plans may be yielding single-digit internal rate of return (IRR) numbers, but considering the benefits of subscriber retention, service differentiation and significant revenue upside potential of data-based services and applications, the strategic value of 3G spectrum is clearly much more than a simple metric-based valuation.

Analysys Mason Limited is a London-headquartered strategy consulting firm in telecom, technology and media.

Sandeep Ladda
Executive Director, PricewaterhouseCoopers

After almost a decade of the first 3G roll-out in developed nations, the Indian government, after much debate and discussion, is on its way to releasing 3G spectrum in the 2100 MHz frequency band.

While the government had set out with an ambitious price tag of Rs 35,000 crore (Rs 350 billion), the question is whether telcos will be able to make profit out of the entire exercise.

The auction of spectrum enabling telcos to provide 3G services is by far the largest auction of any national resource by the government. Also, based on the latest media reports, the auction is expected to generate around Rs 50,000 crore (Rs 500 billion) for the government.

With reduction in the overall fiscal deficit and reduced borrowings figuring high on the agenda of the government, the price discovery (auction) mechanism should fairly meet these objectives.

Telecom companies, on the other hand, face the challenge of serving about 20 million customers who are getting added every month. The total telephone subscriber base has increased to 621 million and, consequentially, the overall teledensity in India has reached 52.74 per cent, as reported by the Telecom Regulatory Authority of India (Trai).

This rapid increase in the customer base is unprecedented and thus poses its own set of challenges. Considering the extremely scarce availability of the basic raw material, i.e. spectrum, and after prolonged delay in its release, telcos are now vying to get hold of any additional spectrum possible.

In the process, for fear of losing out in the race to grab market share, telcos are betting heavily on the pricing of spectrum. Some of them will have to borrow heavily to fund the payout to the government and, subsequently, to fund the capital costs involved in laying down the networks. The heavy interest costs of these borrowings would only increase the cost of acquisition of spectrum.

To compound matters, the current release of spectrum is in the 2,100 MHz band. The spectrum currently held by telcos is in the 800-900 MHz and 1800-1900 MHz bands. The basic law of physics is that an electromagnetic signal of double the frequency travels half the distance.

Thus, the same coverage at 1,800 MHz needs about four times the towers needed at 900 MHz, implying the requirement to deploy new networks, resulting in high capital expenditure, as also acknowledged by the regulators.

Further, there is no "killer application" that would be offered as a result of the launch of 3G services. Most of the current applications like email, chat, social networking, Internet, radio, etc. work well on the current 2G and 2.5G networks.

It is only the "experience" that is better on 3G due to higher data speeds, but there is no "3G only" application that has a mass appeal. Also, it is expected that telcos will initially offer 3G services only in the metros and this may spread across to second tier cities and rural areas over a period of time.

For availing 3G services, customers need to have 3G-enabled handsets that come at a steep cost - another hurdle to convert existing 2G or 2.5G customers to 3G ones. Thus, the challenges for telcos are many.

The aforesaid issues coupled with a continuous drop in the revenue flow because of the tariff war amongst telcos has resulted in a decrease in average revenue per user (Arpu), which is one of the lowest in the world, keeping in perspective the customer base and teledensity.

Another key reason for such low Arpu is that there are 13-14 players in each circle and each telco is getting spectrum in the range of 6-10 MHz, which is too little for telcos to effectively and efficiently utilise the available spectrum to achieve economies of scale.

Summarily, the current bidding process will only result in heftier payouts by telcos. Further lowering of Arpu will make it difficult for them to achieve economies of scale in a reasoanable time frame and thus question the ability to make cash profits as a result of rolling out 3G services pursuant to the spectrum allocation.