Wireless Spectrum Auctions in India: Part2

Let us start Part-2 with an “interesting” topic: Did the operators pay too much for the 3G (and BWA) license? I am not fully qualified to provide a compelling answer, but let us look at some data.

· India’s mobile subscriber base stands at ~600million or 60crores.

· The total cost of 3G would be 67000 crores + cost of rollout. Let us put a ballpark on the cost of rollout, say 50000 crores.

· So the 3G cost/user = 117000/60 = Rs 1950.

· Note here that out of the 600 million subscribers, the number of unique subscribers is much less, maybe 400 million only. This is due to the widespread use of multiple sims in India. This too has to be factored in to the cost. Let us do that by increasing the subscriber cost by 30%. So the cost becomes 1.3x1950 = Rs 2535

· There are variables that may mitigate the per-user cost, such as the increase in subscribers during the 3G rollout phase; however the increase in subscribers can be counter-balanced by an increase in infrastructure cost.

· Since we do not know the cost of BWA spectrum and its associated rollout cost, lets assume that it is 50% of the 3G cost.

· So the cost per user for 3G+BWA would be 1.5x2535 = Rs3800

Let us now look at the Average Revenue Per User (ARPU) in India.

· The ARPU is ~Rs 144 for GSM and ~82 for CDMA users.

· Are all the mobile subscribers equally likely to embrace 3G? Not likely; consider this:

· The Indian market has 95% pre-paid users (ARPU ~Rs 120) and 5% post-paid users (ARPU ~ Rs 550). So revenue-wise the mobile subscribers are very diverse with 5% of the post-paid users providing approximately 5 times the revenue relative to the 95% pre-paid users.

· It is likely that the post-paid users would embrace 3G+BWA. But what about the pre-paid users whose ARPU is the lowest in the world?

· This group is diverse too. It consists of a large chunk of users who use only basic voice services and SMS and may not be interested in any Value-Added-Service that may be provided by 3G. But this group also includes the younger school/college going folks, who at present mostly use voice+SMS, but who may readily embrace VAS such as streaming video and media if given the services at a right price point. The profitability of 3G+BWA would not only be determined by the high paying postpaid users (who constitute only 5% of the subscriber base) but critically by these young sections of the population, who are paying less now for mobile services, but who may pay significantly more if provided the right services at the right price. They would be very important if 3G+BWA is to be profitable in India. Otherwise, given the conservative cost-per-user of ~Rs4000, 3G may drag down the India Telecom sector and lead it to the haemorrhaging way of another much touted sector in India, the Aviation sector. The India government, even in the midst of its present euphoria of collecting a bounty, should be mindful of not letting the Telecom industry go the Aviation way.

Now let us look at another approach to reducing risk and maximizing profitability, the one that many operators have been increasingly demanding. This is mergers & acquisitions. It is a double edged sword. We all know what happened when the telecom licenses in India were allotted in 2008 without any M&A restrictions. Companies like Unitech & Swan, who had absolutely no infrastructure or experience in wireless operations, bought licenses at ~Rs 1500 crores apiece (that too reportedly by raising debt) and sold the licences in a couple of months at ~8 times the value (to Telenor and Etisalat respectively) and became dollar billionaires overnight. And these were not the only ones, there were a handful of others that blatantly profited from the government largesse. This is now known as the Telecom Scam, probably the biggest in terms of the money lost to the exchequer, and the government continues to deny it or at best ignore it.

Thankfully, the recent auctions have restrictions on granting license and on M&A.

1. The winning bidders cannot sell their spectrum before 3 years of granting it and it needs prior approval from DoT.

2. The merged entity cannot have more than 40% market share in any service area, either by subscriber base or revenue.

While these clauses protect consumer interest, they also increase the chances of further tariff wars. The fragmented nature of the industry makes it much more difficult to meaningfully recover the cost of the spectrum and the subsequent service rollout. The government has enabled some reduction in rollout cost by allowing the sharing of network equipment such as:

1. Buildings, towers, fibre

2. Backhaul (wireless and wireline)

3. Even active infrastructure such as antenna, eNodeB, RAN, switches etc excluding the allotted spectrum.

I think the government should monitor the sector closely and should the outlook of the telecom sector deteriorate beyond a threshold, then manage/permit consolidations on a case by case basis. It should also make available more spectrum at a cheaper price in the future and reduce the demand supply gap. The government Memorandum on the 3G+BWA auction mentions the possibility of making available more spectrum; it should honour the pledge should the vital parameters of the industry deteriorate.

I would conclude this section with a short case study on MTNL and BSNL, the two public sector telecos in India.

· MTNL has approximately 6500 crores of cash in its books. In 2009-10, the company actually incurred an operating loss of 823 crores before accounting for taxes and other exceptional items. However MTNL had reported a net profit. This is because its cash balance of 6500 crores earned it an interest income of about 1100 crores. This cash is about to disappear as it has to pay ~6500 crores for its share of the 3G spectrum in Delhi and Mumbai. This would push the company deep into the red despite it operating in 2 of the most lucrative telecom circles in India.

· BSNL too has posted a loss in fiscal 2010. It too has a huge cash balance, reportedly running into 35000 crores. However it too has to pay about 10000 crores as part of the 3G spectrum costs.

However the bottomline is not that these companies are in operating loss. The main concern is that these companies have been losing market share to their private counterparts. The situation is such that they are unable to utilize even 50% of the spectrum and network capacity that they presently have. This is the spectrum, a part of which they now have to buy at a high cost, and also the infrastructure that they have installed at a high cost. This 50% under-utilization of critical resources contrasts with the spectrum and network congestion faced by all the other major operators in India.

What can be a solution? The clean way is to sell/merge these entities to private providers. After-all the government should not be in business (critics may point to China though!). However the government has now chosen a different path that primarily benefits these companies. According to a policy, the GOI may soon allow Mobile Virtual Network Operators or MVNOs. An MVNO can be a

· Thin MVNO, wherein it can purchase the airtime from the parent operator and offer services under its own brand without owning any underlying infrastructure. OR

· Thick MVNO that can rent the spectrum and part of the end-to-end network infrastructure but also provide its own billing, possibly additional infrastructure and its own value differentiation. There can be hybrids that are a mix of the two.

Opening up an MVNO policy would readily benefit MTNL and BSNL as both of them have plenty of spare spectrum and infrastructure. This is a useful way out to make these companies survive without getting into the political and policy minefield of privatizations. So the government may first allow only Thin-MVNOs that only rent airtime and allow MTNL/BSNL to fully own and operate their infrastructure. Then maybe later it would allow a shift to Thick-MVNOs that also put in their own pieces of network components and allow a greater integration of their networks, and then finally (hopefully) slowly allow their integration and merge with these entities.