Not enough spectrum for mobile broadband
Telecom minister Ravi Shankr Prasaad is absolutely right when he takes credit for resolving the issue of the defence band and the defence interest zone since the issue has been pending for years now. It is this resolution, in fact, that resulted in the defence ministry agreeing to swap 15 MHz of valuable 3G or 2100 MHz spectrum for commercial use. Having achieved so much, the minister would have done well to put this up for auction in March instead of leaving this to a later date as he is doing now—as and when the spectrum is available, he has said, he will put it up for auction. It is correct that this spectrum is not available right now, and it take some time for it to be available. But that is the case with a lot of other spectrum as well. To cite the most recent example, the spectrum auctioned last year in February was given to operators only in November and December; indeed in the 3G auction of April 2010, the auction documents even said the spectrum would not be given before October 2010.
More important, the fact is that there simply isn’t enough spectrum in the auction pool right now. The government plans to auction 178 MHz of 900 MHz spectrum, but all of that is already in use today—so, assuming each firm wins back the spectrum it has right now, this doesn’t create fresh capacity. There is 73 MHz of new spectrum—and 26 of old spectrum—in the 1800 MHz band but since much of this does not comprise contiguous chunks of 5 MHz each, it cannot be used for offering data services. Ditto for the 104 MHz of the 800 MHz band spectrum; that leaves only 85 MHz of 2100 MHz (or 3G) spectrum that is really a solid addition to the pool that can be used for data services.
Two issues result from this; the first is capacity and the second is cost, and both are inter-related. Take capacity first. Taking low usage assumptions of 600-700 MB per month, each 3G slot of 5 MHz can accommodate roughly 50 million users—so, as compared to the 65 million 3G subscribers that India has at the moment, the three 3G slots allotted to private telcos can deliver a maximum of 150 million subcribers—neither MTNL nor BSNL have done much with their 3G spectrum. If the usage doubles, as it should in a proper broadband scenario, this number will halve. To put this in perspective, India’s target is to have 600 million mobile broadband subscribers by 2020.
Releasing more slots of 3G spectrum later will theoretically help achieve the target, but this is where the issue of costs come in. With a debt of over R2 lakh crore—even if you exclude what Bharti Airtel took to buy Zain in Africa—there is just that much more the telecom industry can borrow to buy more spectrum; in other words, it is certain there will be some failed auctions over the next few years. FE’s Spectrum First series had done some calculations for FY16 and found that while telcos would need to spend $12-13 billion in interest costs alone if they were to buy the spectrum likely to be on auction (goo.gl/CyfnBG), their ebitda is likely to be just around $6-8 billion. So, with spectrum costs being driven sky-high in the intervening period, only the very deep-pocketed firms will survive which, in turn, will mean broadband costs will remain high and will, in turn, restrict usage.