|Telecom wars, part x|
|Thursday, 23 March 2017 04:29|
This time around, it is the govt that will also bleed, and it has only itself to blame
With telcos like Bharti Airtel or Vodafone-Idea at 10-15% of their data capacity, few are going to buy outrageously priced spectrum for the next few years — the days of easy auction-money are over
Apart from the cost synergies worth $10bn (on an NPV basis) that Vodafone and Idea are expecting to extract from their merger, perhaps the biggest benefit will be the fact that, for a year or two at least, the combined entity doesn’t really need to buy any more spectrum. There is the obvious increased capacity it gets by merging the Vodafone and Idea holdings but, equally important, it has at least 80% of its data capacity unutilized. Given Bharti Airtel and RJio, the other two big players who now dominate the industry, also do not need to buy more spectrum immediately, the biggest sufferer in the next round of the telecom wars will be the government.
While the industry’s revenues will contract in FY17, as then telecom secretary JS Deepak pointed out to Trai a few days before he was moved out, worse lies ahead. Though instalments worth Rs 3 lakh crore will be paid out over the next 11 years for the older auctions, keep in mind fresh auctions generated around Rs 2 lakh crore in the last seven years – so keeping up a supply of fresh auctions is critical. Between FY11 and FY17, 43% of government revenues from the sector came by way of auction fees and, with industry revenue stagnating, even the annual license/spectrum fee revenues will stagnate.
The maths of why another auction won’t happen for a while is simple. On average, based on their current spectrum holdings, telcos get around 50 GB of data capacity per day per base station for every 5 MHz of 3G/4G spectrum they hold. So, for a Vodafone-Idea which has around 30MHz of such spectrum on a pan-India basis and around 200,000 towers (not all are 3G/4G, but converting them is relatively cost-effective), this means a data capacity of around 22 billion GB per year. Since the current usage is just around 1-2 billion GB, the combine doesn’t need more data spectrum till its usage rises at 10-12 times, either by way of new customers or by way of each one using more.
The capacity utilization numbers aren’t significantly different in the case of Bharti Airtel. In the case of RJio, its free service has ensured most of its 14 billion GB capacity is being utilized, but once it starts charging for data next month, the 12-15 GB being used per customer every month could fall significantly – also, once it adds more towers from the current 125,000 or so, it can increase its data capacity without needing to buy more spectrum.
Were spectrum to be priced reasonably, it is true, the financially strapped industry – the best-performer Bharti Airtel has a leverage of around 2.8 times – could consider buying more of it despite debt levels of Rs 4.6 lakh crore (that’s 2.6 times its adjusted revenues for FY17). But that requires a sea-change in the government’s rapacious attitude – compared with 11% in FY07, the government took away 32.4% of telco top line in FY17 by way of recurring licence/spectrum charges and auction costs; this goes up to an even higher 47.4% if service taxes are included.
So, in the case of the 3G auction for 2100MHz spectrum in 2010, a peculiar auction structure and the fact that eight telcos were chasing three pan-India slots meant the price for Delhi and Mumbai rose to 10 times the reserve price – while the two circles accounted for 13.7% of industry revenue, their spectrum price was 40% of that year’s auction.
It continued to get worse. In the 1800MHz auction in 2012, the reserve price was set almost equal to the bid price in the 3G auction in 2010. In Delhi, while the 2010 auction had a reserve price of Rs 64 crore per MHz and an auction bid of Rs 663 crore for 2100MHz, the 1800MHz reserve price in 2012 was a whopping Rs 554.4 crore. The telcos who bought the spectrum had no plans to offer 3G services but had to buy it for the inferior 2G services as they had no spectrum.
In 2015, with their 900MHz licenses expiring, telcos had no option but to bid sky-high rates – this time around, the government set a reserve price 3-4 times that of the 3G price in 2010. In Maharashtra, the bid price was Rs 251.6 crore per MHz in the 2100MHz band in 2010; the reserve price for 900MHz in 2015 was kept at Rs 525 crore. Once again, the spectrum on offer was very low and, with no attempt to increase supplies by ‘harmonisation’ of existing spectrum, prices sky-rocketed. There was no rationale for this since the comparable 800MHz band was priced much lower but telcos couldn’t bid for this as there was no 3G equipment in this band.
While reserve prices have been dropped due to lack on demand on a few occasions, as a general rule, these have generally been based on the 3G auction in 2010. As a result, the industry has spent around Rs 3.6 lakh crore in buying spectrum since 2010. Though Deepak’s note to Trai hinted at the need to rationalize rates, he was probably too mild and, in any case, it is unlikely Trai is going to come up with a sensible lowering of reserve prices under its current chairman – keep in mind, the last auction for 700MHz spectrum failed because, completely arbitrarily, Trai fixed its reserve price at four times the last auction bid for 1800MHz spectrum.
Also, given the surcharged business-equals-corruption atmosphere in the country and the plethora of PIL-activists ready to challenge anything in court, any sensible pricing will be tagged suit-boot-ki-sarkaar. The saving grace is that, with government revenues in trouble, the government has no option but to get its act together, sooner rather than later.