Analyst Angle: India spectrum auction likely to be aggressive, yet realistic, affair: Qualcomm's bid for BWA spectrum threatens WiMAX prospects in India

After several postponements, months of very public inter-ministerial bickering that required the intervention of a special task force of senior ministers (the so-called Empowered Group of Ministers) headed by the finance minister, and a few lawsuits, it’s finally on. India finally begins today the public auction of spectrum for third generation mobile telephony and, subsequently, of spectrum for broadband wireless access.
Given that the number of bidders far exceed the number of available spectrum slots, and the fact that outcome of the auctions will likely trigger a shakeout in the rather crowded Indian telecom market, the bidding process is likely to be fairly long drawn – and aggressive.
However, the aggressiveness is likely to be tempered by market realities that remain in India and by the lessons of history. India is predominantly a prepaid, voice-oriented market with as-yet unclear data revenue prospects. The culture of mobile data consumption, in the absence of an established and locally relevant eco-system, will likely take some time to set in and, in the early stages, leading mobile operators may well use the 3G spectrum to address the capacity crunch they face on their current 2G networks, especially in key metropolitan centers. (Recall, we regularly hear about millions of new mobile subscribers coming on to the networks in India, but rarely, if ever, of any corresponding growth in the availability of spectrum!)
Avoiding the ‘winner’s curse’
Further, one can assume that Indian players will have learned from the mistakes of some of their western counterparts who, failing to recognize the many challenges of monetizing 3G networks and services, had gone somewhat overboard in their spectrum bids earlier in this decade. Nobody wants to suffer the “winner’s curse” after all. Remember NextWave?
The design of the auction process itself is also likely to mitigate against irrational exuberance on the part of bidders. The auction process is designed to deter predatory bidding; only the auctioneer is allowed to raise the bids – and that, too, in a pre-determined manner, using a publicly stated formula – and never the bidding players. Further, the process is also designed to thwart speculation; the players must make the full payment for their bids within 10 days of winning the spectrum – or risk losing their millions in earnest deposits necessary for participation in the auctions.
That said, it is safe to assume that the final price will likely be driven by the “Nth plus 1” player – that is, by the final player bowing out of the race.
Some background notes
But, perhaps, before I proceed further, a little background may be in order – if only for the sake of those who may have been too preoccupied with other pressing existential issues to pay attention to the Indian scene.
India is divided into 22 mobile telephony operating areas, commonly referred to as circles. However, not all circles are equal; they are stratified into A, B and C categories, depending on population and, by inference, revenue prospects. Each of the country’s four leading metros – Delhi, Mumbai, Kolkata and Chennai – for instance, is designated as an A circle, as are a few states in the country.
Each of the 22 circles is simultaneously up for bids in the current spectrum auction. The auction is designed as a multi-round online process in which the auctioneers – NM Rothschild & Sons and dot.econ – present bids in an ascending order. The bidders can accept or decline the auctioneer’s proposed price, but cannot offer their own bids. Each player must bid in a given round to preserve the right to bid in a subsequent round.
It might be opportune to reiterate that there are two separate, if related, auctions – one for 3G spectrum and the other for BWA (broadband wireless access) spectrum. The auction that starts today is for 3G spectrum. The BWA auction is scheduled to start two days after the completion of the 3G auction, whenever that may be. The two auctions follow the same set of guidelines.
Spectrum slots and bidders
For 3G, the Indian government is auctioning three paired spectrum slots in each of 17 circles, and four slots in the remaining five circles. Each slot is 2X5 megahertz in 2.1 GHz band. For BWA, the government is auctioning two unpaired spectrum blocks of 20 megahertz each in 2.3 GHz band in each of the 22 circles. The Indian government is impartial to the choice of BWA technology.
The base price for each pan-India 3G slot has been pegged at Rs 3500 crores (roughly $780 million) while that for each pan-India BWA spectrum block has been set at half that amount, Rs. 1750 crores.
The winners will be awarded spectrum in September, when the Department of Defense vacates the spectrum.
There are nine bidders for the three/four 3G slots, and 11 bidders for the two BWA slots.
In the 3G case, six of the nine players – Aircel, Bharti-Airtel, Idea Cellular, Reliance Telecom, Tata Teleservices, and Vodafone-Essar – have signaled their intent to bid for all 22 circles. The other three players – Etisalat, S Tel and VideoCon – have signaled only regional ambitions for 3G.
In the BWA case, there are, surprisingly, 11 players in the race. Eight of them – Aircel, Bharti-Airtel, Idea Cellular, Infotel Broadband Services, Qualcomm Inc., Reliance WiMAX, Tata Communications and Vodafone Essar – have indicated an interest in seeking pan-India spectrum position. Those seeking regional BWA footprints are Augere (Mauritius), Spice Internet Service and Tikona Digital Networks.
The rollout obligations are fairly liberal, by most standards. Spectrum winners, or licensees, must provide 90% coverage in leading metros, and 50% coverage in other A, B and C circles, within five years of the day they are granted the spectrum, or face punitive damages that could include eventual withdrawal of the right to the spectrum.
Special treatment for state-owned incumbents?
Incidentally, the spectrum slots being auctioned are in addition to what state-owned incumbent BSNL and its sister organization, MTNL, have already been allocated for 3G and BWA. The state-owned BSNL/MTNL combine was allocated the spectrum more than a year ago to give them a head start, with the understanding that they would pay the price paid by the highest bidder in the auction process.
Following their till-date dismal performance, the government is now considering ways – to the understandable chagrin of commercial players – to provide relief to the ailing incumbents. Among other things, the government is considering giving the incumbents an additional 24 months to pay for the spectrum.
Qualcomm’s play
Qualcomm’s decision to bid for the BWA spectrum has upset the WiMAX apple cart in India, threatening its prospects in the country. What looked like a sure thing for the WIMAX contingent in India, now looks anything but. WiMAX proponents had assumed that they would have a relatively easy and uncontested run of the Indian market, since 2.3 GHz band seemed to have been globally harmonized for WiMAX. Further, nationwide deployments in India would finally provide the necessary scale for the technology that could be leveraged globally. Oh, well!
Qualcomm has said that if successful in winning pan-India BWA spectrum, it intends to secure Indian partners to build out a TD-LTE (Time Division-Long Term Evolution) network that will, eventually, be spun off to suitable players in the given market.
The move is classic Qualcomm, in the mold of their strategy with CDMA and FLO-TV – invest in a fledgling technology in which it has a stake, nurture and help accelerate its adoption and then, once the eco-system is in place, to spin off the operational unit and rev
ert back to its core business, of serving the market it helped creat
e through chipsets. Qualcomm has a large portfolio of OFDMA patents – augmented by its 2005 acquisition of Flarion Technologies – that are relevant to TD-LTE.
Qualcomm’s entry in the Indian BWA spectrum auction may have sparked a WiMAX/TD-LTE war not unlike the GSM/CDMA wars of the past. For Qualcomm’s move, unsurprisingly, prompted WiMAX proponents to quickly allege that the San Diego-based chipmaker was planning to “hoard” spectrum and slow broadband growth in the country. The WiMAX folks argue that TD-LTE will not be ready for a few more years and the delay in BWA rollout will thwart national ambitions and policy goals. Some in the media have echoed similar concerns, stating that delayed BWA rollouts may not be in the public interest.
Disingenuous argument and fungibility of spectrum
The allegations against Qualcomm seem somewhat disingenuous; worse, they demonstrate a poor understanding of the spectrum situation in India, and underestimate the possible fungibility of spectrum between 3G and BWA in the short to medium term. Spectrum is, after all, the lifeblood of all that is wireless and mobile telephony and, as they say, one can never be thin enough, rich enough, or have enough spectrum.
Even a cursory look at the list of BWA spectrum bidders makes it clear that many of the leading players seeking 3G spectrum who are also seeking BWA spectrum have indicated little, if any, interest in deploying WiMAX. The reason for this, probably, is that mobile operators – given the meager 5-megahertz paired spectrum being auctioned at present — are seeking to ensure they have enough spectrum for future growth, when the LTE technology becomes more viable.
Qualcomm’s logic for getting into the BWA fray seems to be a variation on the same theme. The company’s India head has stated that the decision to bid for the BWA spectrum was prompted by the desire to “de-risk” 3G deployment for mobile operators by ensuring that these operators have both spectrum and a migration path in the future.
Further, while one agrees with the view that delayed BWA rollout would be inimical to the public interest, it might be over-reaching to hold that WiMAX deployment may be the only answer. Sure, there is a case to be made for WIMAX, but it is an open question if the mobile variant is as ready, or the necessary eco-system as mature, as WIMAX cheerleaders would have us believe.
In any case, the rollout obligations have stipulated a five-year time frame. It is unlikely that whosoever wins the BWA spectrum would default on these obligations. As a matter of fact, it may well be argued that competitive pressures and enlightened self-interest will likely drive the winner – whether a WIMAX or a TD-LTE backer – to beat the stipulated network rollout deadlines.
Further, as I have argued elsewhere, new technologies are rarely born in a vacuum; rather, they are brought forward in existing structures of power represented by legacy systems. These legacy systems represent not merely a set of technologies and networking gear, but also relationships among industry players, between industry players and capital markets, and between technologies and their developers.
In other words, they represent working ecosystems or ecologies of support that remain in place for long periods of time and continuously evolve with newer technologies. It is rather difficult for any new technology to supplant a legacy system, its ubiquity and backward compatibility (as well as its potential trajectory) unless it is orders of magnitude more valuable to the industry and, more importantly, the end customer. That case for WiMAX has yet to be made.
Squandered opportunity?
But beyond issues relating to lack of ubiquity and backward compatibility, there are other challenges WiMAX must address. Key among these might be that, for all its merits, the technology seems to be losing support of vendors that could have been critical to its global success. For instance, after Nokia-Siemens and Alcatel-Lucent, Cisco seems to have muted its support for WiMAX access infrastructure that it had signaled with the 2007 acquisition of Navini Networks. Cisco may not have withdrawn its support but, following the acquisition of Starent Networks, the vendor will likely shift its support from access to the core network.
The perceived waning of support from leading infrastructure vendors may translate into a squandered opportunity in India (and elsewhere) for WiMAX. But, that said, I believe it may be a mistake to read Qualcomm’s entry into the BWA auction as necessarily the beginning of the end for WiMAX in India. BSNL, Tatas and Reliance, for instance, are still firm in their support of nationwide WiMAX deployment. The competitive threat from TD-LTE might just be the necessary fillip that WiMAX may have needed all these days.
BWA spectrum bidding could be fierce
Meanwhile, for the reasons outlined above — particularly, the fungibility of spectrum over time — it might be reasonable to argue that bidding in the BWA spectrum auction (that follows the 3G spectrum auction) could be relatively fierce, since those unable to acquire the 3G spectrum slots are likely to make a stronger push for the BWA spectrum.
Bharti-Airtel and Vodafone are the odds-on favorites to win 3G spectrum in most of the circles, with Reliance a close third. The aggressive bidding, even as it is tempered by realities of the Indian market, is likely to stretch the balance sheets of leading mobile operators in India. Regardless of who wins or loses the spectrum bids, there is one party assured of a reasonably happy outcome in the short term – the Indian Treasury. The government is likely to earn anywhere between $8 billion and $ 10 billion from the two auctions.
Let the games begin!
Dr. Shiv Bakhshi is founder & principal analyst of Mobile Perspectives. He can be reached at shiv@mobileperspectives.com