|No more phones for Bharat|
|Thursday, 26 April 2012 00:08|
Trai refarming proposal makes rural telephony unviable and BSNL/MTNL will have to fork out Rs 27,000 crore
After all the huffing and puffing over the Trai’s latest recommendations, if they only result in phone tariffs rising 3.6 paise per minute, what’s the big deal? And what’s wrong with the proposal on ‘refarming’ spectrum—taking away more efficient spectrum in the 900 MHz band and replacing it with 1800 MHz spectrum? After all, argued a former boss who I called to rant about his newspaper’s editorial supporting Trai, the older telcos who have the 900 MHz spectrum don’t have a God-given right to keep it for the rest of their lives. These, and a few more such, are the gist of the arguments made by those who think Trai’s done the right thing.
While the older telcos—a third, the largest chunk of 900 MHz spectrum, is with PSUs BSNL and MTNL—argue their licence conditions actually bind the government to renew their 900 MHz spectrum when the current licences expire, let’s concede the argument for now. At a basic level, the argument of a ‘level playing field’ where everyone gets the right to use the more efficient spectrum is an appealing one, never mind that newcomers always bear a higher risk and the price of this is a first-mover advantage—Tata Steel getting better iron ore mines than a Jindal Steel, for instance. When you consider the government stands to earn R7 lakh crore from the Trai recommendations, assuming just the reserve price for spectrum, the case in favour of Trai gets even stronger.
Problem is the facts aren’t quite what they’re made out to be. For one, as Table 1 (derived from Trai’s Annexure VI) shows, India’s reserve prices are on average 70-80 times more than those in countries who have many times India’s per capita income. But if that’s true, how can the tariff hikes be under 10% as the Trai argues? What Trai does, in Annexure VII, is to look at the total amount of spectrum telcos have, multiply this by the Trai reserve price, calculate the EMI of this … and, voila, you get a 3.6 paise hike. Problem is Trai’s maths doesn’t quite add up. Trai does its calculations based on 576.2 MHz of spectrum with the industry but its own report says 1160 MHz has been given out in various 2G bands—just based on this, Trai’s 3.6 paise becomes 7.2 paise, or a 16% hike. In Table 2, I do the calculation in a different manner which gives a much higher hike, but those who get freaked out by maths would do well to skip it. Given India has the lowest tariffs in the world, surely a 17% hike is peanuts?
Trai and all its followers will have us believe ‘refarming’ is revenue-neutral since telcos will just give up the 900 MHz spectrum and get 1800 MHz spectrum in return. This is misleading since there are huge costs involved. For one, as compared to the R1,650 crore paid in the 2001 auctions for 6.2 MHz of spectrum, telcos have to pay a minimum of R3,622 crore per MHz of new spectrum in the 1800 MHz band. Since these have to be paid in 2017 for MTNL and 2020 for BSNL, you apply interest rates to come to the one-time charge at that time, and then discount it to arrive at today’s value of this additional cost—turns out this is R27,000 crore for BSNL/MTNL, a little more than Bharti Airtel’s R26,000 crore.
There is then the capex since, for rural India especially, the number of telecom towers that need to be put up in the 1800 MHz band is roughly double that in the 900 MHz band (that’s why 900 MHz is preferred)—this doesn’t apply as much to cities since the density of population is so large there are a lot more towers in them. In the case of BSNL, adding the additional towers could add anywhere near R15,000-20,000 crore.
Even this, of course, is chump change compared to the R7 lakh crore the government can get, so it may still be worth it. Except, there’s one problem: once you replace 900 MHz spectrum with 1800 MHz spectrum, the case for rural telephony pretty much disappears. In which case, rural telephony’s spread will either slow down or some explicit subsidies will have to be given. There’s no escaping the maths this time around (Table 3), but I’ve tried to keep it simple. On average, one Base Transceiver Station (BTS) serves around 1,000 customers—given the revenues (R100 per month per customer) and the costs of running the BTS, you still earn R15,000 per BTS. Since you have to set up roughly double the number of BTS’ once you’re in the 1800 MHz band, the revenues halve and you get a loss of R22,500 per month per BTS.
In which case, whether it’s bye bye Bharti, Vodafone or BSNL/MTNL or not, it’s certainly bye bye Bharat as far as telephony is concerned.