Auction of 3G spectrum in India

Preface: If you are telecom professional, this article is for you. If you are not working in telecommunications domain and want to know for personal interest about regulatory happenings, pricing decisions of spectrum or want to know why politicians have affinity towards telecommunications ministry portfolio, you can read on. For rest of you, it can be is dry, boring!

3G spectrum auction

With mobile phone industry adding ~10 million subscribers consistently every month, it has become a spotlight of modern Indian expansion. To the media, it is a way to express and show the world how dramatic the economic scene is changing in India. Barely 15 years back, people could think of reaching their near ones located in other towns only by placing trunk call requests with post office and had to wait for a callback to get connected. This apart from hefty, often scary prices of making long-distance calls! This compared to today… where enterprises, SMEs, individuals can make local & long-distance calls for affordable prices has in a way changed the way business is done. The multiplier effect of mobile growth has altered social fiber, media has used this to show to outside world with media-multiplier... on how students have hooked on to messaging… on how pan wala, construction worker and common man on street have changed way they communicate and go about their jobs.

In India, DoT is arm of government which controls the allocation of spectrum. Strategy and pricing is done in close inter working with telecom ministry. The commercial spectrum is allocated to 2G licencees from 1996. The rest of spectrum is large and closely guarded and used for military purposes. The allocated spectrum to large commercial players has been used up. There is an upper limit to the number of subscribers which can be serviced by allocated spectrum to mobile service providers. With number of subscribers rising exponentially, mobile operators need more spectrum to maintain same quality of service. Although mobile technology has been long researched and has now reached a point where it is wide commercial deployment, it is spectrum which is scare resource blocking launch of newer services. For 3G and BWA services, the additional spectrum can come by way of using existing spectrum in a better way or by allocating spectrum meant for the military. It is no surprise that since spectrum is scarce, by demand and supply theory its price is soaring. It is also not surprising that real estate players are entering business of buying spectrum license in hope that they can sell it for higher price later (like real estate).

Mr.Pranab Mukherjee has recently increased estimated value of 3G/BWA auctions from Rs.20,000 crores to Rs.35,000 crores in latest budget. Since then, finance and telecom ministries are debating reserve price for 3G auction. Telecom ministry wanted reserve price to be Rs.2,020 crores whereas finance ministry wanted it to be Rs.4040 crores. The prime minister has reacted firmly by setting up GoM (group of ministers) to decide on the issue. The GoM will create checks for telecom ministry which has been under fire for several questionable decisions in past. The GoM is a welcome move to ensure quick decisions, especially issues like annual revenue-share licenses have to be decided. It will also decide on how many number of revenue slots will be recognized etc. GoM would do well by keeping couple of things in mind.

* Keeping lower revenue sharing for 3G is not great idea – we know how few large well known firms are being investigated for alleged disguising earnings to fall into lower revenue slab.

* Distinction between 3G and BWA services is very thin. Voice traffic also can be transported as data. Once a firm gets BWA license, there will be enough devices in market to provide voice services on the same spectrum.

Why would telecom ministry want to have different reserve price for 3G and BWA services? (Proposed rates for 10 MHz of 3G is Rs.2,020 while for 20 MHz of BWA is Rs.1,010). The lower rate for BWA is justified since BWA devices cannot be used seamlessly for voice. However since number of phones sold are much more than computers sold, more and more of broadband traffic will get diverted from computer to mobile devices. A faulty government policy prevents providing voice services on BWA platform since it doesn’t allow full fledged internet telephony. However in future, when government allows internet telephony completely, BWA licensees will have additional voice capability for no additional costs! They can terminate calls to landline and other mobile phones over BWA spectrum which was bought for lower costs during this auction.

A peek into history - 3G spectrum auction at Britain in 2000

$34 billion was raised by British government for auctioning 3G telecom licenses which concluded in Apr 2000! It was biggest ever auction (till then) in history. The amount of money obtained by selling air to 5 telecom companies raised eyebrows and also doubts in many economists. Till then, there were 4 2G operators viz. Cellnet, One-2-One, Orange and Vodafone. The state owned British telecom eventually bought Cellnet after it was privatized by Margaret Thatcher. Britain had seen an exponential rise in use of cellular phones and was eager to see high speed data, internet access on mobile phones. 2G licenses had been sold by way of beauty contests involving a published set of criterion; not strictly through an auction. It was only after US successfully auctioned off 2G licenses in 1994 did it want to try same for 3G licenses. Rather than selected few bureaucrats judge a beauty contest, auctions were a better method to allocate resources to businessmen who could put their prized bids to best use.

The most common fear in auctions is that end consumers like us will be pushed with higher prices. Like all businesses, telecom companies will charge prices that maximize their profits while keeping competing in mind. At this stage, telecom companies cannot gauge how much an end consumer would be willing to pay for 3G services. Consequently there is a tendency that auctions will be hard fought tending to keep prices low. However during the auctions, bidders will be forced to increase prices in view of competition. Here is where governments should be well aware of ceiling prices; which will ensure that telecom companies survive at end of ordeal and continue to fuel growth of economy in multiple ways.

The published aims of auction were

* To assign spectrum efficiently

* To promote competition

* To ‘realize the full economic value’

Like in India, there were similar regulations that no bidder would be allowed to own more than 1 license and no winner would be allowed to own a license and not put it into commercial operation for more than a year. One of issues which auctioneers felt is that incumbent 2G operators had advantages over new entrants due to established consumer base, brand value and cross optimization possibilities with existing 2G infrastructures. The incumbent operators would be willing to shell out more money for same air. However it will not satisfy 3^rd objective of government to realize full economic value by having a set of colluded incumbent bidders. So like in India, they also mandated ‘roaming’, which would allow an entrant access to incumbents’ network at a regulated price. Another concern of auctioneers was to decide on number of licenses. It is an established theory that 4 players can profitably operate in any business (5^th,6^th and more players erode profit levels, eventually destroying incumbent players). Again like TRAI did in India earlier, they decided to restrict number of licenses to 5. However telecom ministry is now awarding more than 6 licenses in different circles. Bundling of spectrum in various packages was also tried:

* License A – 2 X 15 MHz of paired spectrum plus 5 MHz of unpaired spectrum

* License B – 2 X 15 MHz of paired spectrum with no unpaired spectrum

* License C,D & E – 2 X 110 MHz of paired plus 5 MHz of unpaired spectrum

Incumbent operators were allowed to bid only for B, C, D and E licenses where A was reserved for new operators. 2 A-licenses were on the tender, ensuring that one of them would go to a new operator. An Anglo Dutch auction was used to find best bids (refer to wikipedia for more details on what is Anglo Dutch auction).

The auction design ensured that if new players had only been interested in the 2 large licenses, the competition would have spilt over to smaller licenses also. There were many similar designs within auction which helped government to provide structure to high revenue event. E.g. Incumbent operators who could try to prevent ‘roaming’ were strategically prevented from doing so during the auction. One loophole was that bidders were not asked for a large enough corpus security deposit. Bidders didn’t feel heat of pulling out of race mid-way which could tactfully lure competitors to put in more money than workable price.

It is another issue that this auction worth $34 billion is believed to have led to telecom bubble bust in 2002!

Closing words

There is a clear need for an auction design in India. Auctions are not a matter of ‘one size fits all’ but is more ‘horses for courses’ where auction design needs to be elaborated based on current situation. In India too, a repeat of British 3G auction is not desirable. A well crafted auction is required, not just based on plain economic gains. The result of auction should create an environment where all players work towards creating efficient data services rather than all players limping in economic debts.