Making the straightforward complicated Indian-style – reprise

Citation

Curwen, P. (2010), "Making the straightforward complicated Indian-style – reprise", info, Vol. 12 No. 6. https://doi.org/10.1108/info.2010.27212fab.001

Making the straightforward complicated Indian-style – reprise

Article Type: Rearview From: info, Volume 12, Issue 6

A regular column on the information industries

A Rearview published in this journal in 2009 (Vol. 11 No. 4) under the same title (bar “reprise”) explained the circumstances surrounding the restructuring of the Indian mobile market and the stumbling progress towards the issuing of 3G licences from the perspective of end-December 2008. At that time, the (several times postponed) auction rescheduled for January 2009 was about to be postponed yet again. This auction has now finally taken place in June 2010, and the past 18 months have witnessed sufficient changes in the circumstances of the Indian mobile sector that the story warrants an update.

In January 2010, it was finally decided to auction three blocks of 5 MHz paired in all 22 “circles” in addition to the licences previously granted to BSNL and MTNL, which aggregated to one licence per circle as noted in the previous Rearview. This decision was later amended to four blocks in respect of five circles, making a total of 71 licences on offer.

Largely as a result of Ministry of Defence-owned spectrum not being passed over as promised, the auction winners were not guaranteed to receive their spectrum immediately after the auction was completed, but they were nevertheless required to make a 25 percent deposit immediately and to pay the rest after 15 days. It was also decided that the licence fees could be paid in rupees and refinanced by overseas borrowing within one year of paying the final instalment. However, in February 2010, the Ministry of Law and Justice warned that if the government auctioned spectrum that it did not own, the operators that won it might be able to ask for their money back plus interest.

There were also changes proposed in respect of usage charges. At the time, 2G operators paid between 2 per cent and 6 per cent of annual gross revenue (AGR) depending on the spectrum held. The regulator (TRAI) tabled a proposal that recommended a further 1 per cent for operators with more than 6.2 MHz of spectrum. For its part, the Ministry of Communications and Information Technology proposed to the Empowered Group – effectively acting as adjudicators – that GSM operators with up to 8.2 MHz and CDMA operators with up to 5 MHz should pay an extra 1 per cent while those with more spectrum should pay an extra 2 per cent. This meant that a 2G operator gaining a 3G licence would end up paying between 3 per cent and 8 per cent of AGR.

Despite the financial and legal uncertainties introduced by the above developments, applications were invited for delivery by 19 March with qualified bidders to be announced on 30 March and bidding to commence on 9 April. Nine bidders were deemed to have qualified: Aircel, Bharti, Etisalat, Idea, Reliance Communications, S-Tel, Tata, Vodafone and Videocon Telecoms. Uninor (controlled by Telenor) and Loop declined to participate. It was unclear whether any individual operator would attempt to acquire a licence in all 22 circles, thereby guaranteeing pan-Indian 3G coverage – most did not anyway intend to provide it for 2G – but the cost was expected to deter this.

The bidding initially concentrated upon the popular circles such as Delhi, Gujarat and Mumbai, but by day three the cost of acquiring a pan-Indian licence was already equivalent to $973 million and the government announced that it might refund a significant percentage of what BSNL would have to pay (to match the highest private bids) in order to stabilise its finances. The $1 billion dollar mark was passed for a notional pan-Indian licence on day four; the $2 billion dollar mark on day 17 (after 100 rounds of bidding); the $2.5 billion mark on day 23; and the $3 billion mark on day 26. The auction ended on day 34 after 183 rounds with bids amounting to $14.63 billion, including the matching prices required of BSNL and MTNL. As expected, no pan-Indian licence (which would have cost $3.6 billion) was acquired. Aircel won in 13 circles at a cost of $1.4 billion; Bharti Airtel in 13 circles at a cost of $2.66 billion; Reliance in 13 circles at a cost of $1.86 billion; Idea in 11 circles at a cost of $1.25 billion; Vodafone Essar in nine circles at a cost of $1.74 billion; and Tata Teleservices in nine circles at a cost of $1.34 billion. S-Tel won in three circles but Etisalat and Videocon were left empty-handed.

Almost immediately upon the end of the 3G auction there took place a so-called BWA auction, which was effectively for the provision of 3.9G/4G services although, in principle, these could be provided using LTE over 2G/3G spectrum. There were 11 qualified bidders: Aircel, Augere, Bharti, Idea, Infotel Broadband Services, Qualcomm, Reliance WiMax, Spice ISP, Tata Communications Internet Services, Tikona Digital Networks and Vodafone. The auction commenced on 23 May with the price for the two pan-Indian licences set at a minimum of $367 million apiece. The auction ended on day 16 after 117 rounds with the cost per notional pan-Indian licence reaching $2.74 billion. Infotel Broadband Services acquired a pan-Indian licence and declared its intention to use WiMAX. Bharti won in four circles at a cost of $695 million and Aircel in eight circles at a cost of $723 million, while Tikona Digital Networks won in five circles and Augere won in one circle. The most interesting aspect was that Qualcomm won in four circles, including Delhi and Mumbai at a cost of $1.05 billion, and declared its intention to use TD-LTE even though the spectrum being used for this new technology in China was different to that in India so no kit or devices were as yet available. A total of $5.48 billion was raised, rising to $8.23 billion when the matching BSNL/MTNL licence was added in.

Subsequently, BSNL and MTNL requested that their 3G/BWA fees be refunded on the curious grounds that they had not had any say in whether they would be awarded their licences. The regulator, predictably, proved to be unsympathetic.

The above narrative begs a number of questions, the first of which concerns the future prospects for 3G in India. The experience of BSNL and MTNL prior to the auction had been less than promising. In late May 2010, MTNL announced that it had acquired a total of 400,000 3G subscribers, representing less than 10 per cent of its total subscribers despite the fact that, in Delhi, for example, 2G subscribers were being offered “free 3G” with unchanged voice charges plus charges for data used if upgrading to 3G. But most handsets are not 3G-capable so more subsidies may be needed. This is hardly a scenario that will appeal to operators that ended up paying considerably more than they had hoped for their 3G licences, especially given the increasingly fierce competition in every circle that had already driven down per-minute charges to a barely profitable level.

To some analysts, the 3G auction was reminiscent of the over-exuberance manifested during the auctions in Germany and the UK ten years previously, and clearly much the same issues applied: in particular, could one operator in a circle afford to be 2G-only if other operators were providing 3G? Only Uninor claims to believe that this makes economic sense, but the 3G licensees must now devise a strategy that will earn them a return on their fees. On the plus side is the ready availability of W-CDMA and cdma2000 1xEV-DO kit and devices, but the latter may well need to be subsidised. A second issue is whether, and when, the spectrum will become available, and how it will be rationed if there is not enough to go around.

The other issue of considerable interest is the implication of the auctions for the structure of the Indian mobile market. As noted, the main private 2G operators played a modest role in the BWA auction, perhaps because they had overspent for their 3G licences, and hence it provided the opportunity to introduce even more competition in the provision of data services. What was especially notable was that Reliance Industries (which had but recently agreed to discontinue a non-compete clause in relation to Reliance Communications) immediately agreed to buy 95 percent of Infotel for $1 billion.

As for the 2G/3G structure, it has to be said that the use of circles in India is virtually unique – there are some similarities in, for example, the USA and Russia, but both of these have a handful of national as well as large numbers of regional operators. With 450 million subscribers, second only to China, India can clearly sustain more than the four or five operators typically present in developed markets, but there are currently more than ten, which seems excessive. Economic logic suggests that consolidation must take place, especially between operators with minor overlaps in coverage, but the most obvious recent trend is for foreign companies to buy stakes, and this has yet to run its course since Reliance Communications, for example, is currently looking for a buyer for a 25 percent stake to help pay for its licences. The key problem in mergers or in setting up joint ventures is always the need for agreement on the value of the two networks involved, and this is likely to be a fraught process since 3G services have yet to be launched.

Clearly, therefore, this multi-year saga has yet to run its full course. Is a reprise of this reprise on the cards?

Peter CurwenVisiting Professor of Telecommunications at the Department of Management Science, Strathclyde University, Glasgow, UK.