Former Comptroller and Auditor General Vinod Rai’s parting kick to Messrs Kapil Sibal and the UPA on spectrum is proving to be much stronger that his initial, mind-numbing allegation that Andimuthu Raja’s 2G scam caused a loss of Rs 1,76,000 crore to the exchequer.

To be fair to Rai, his 2G spectrum report made several calculations of presumptive loss, but it was the highest number that the media went to town over, thus making it appear as though this was the final authoritative figure.

It wasn’t. Prices and presumptive losses depend on demand, supply and market sentiment. We can presume some loss, given special circumstances, but exact calculations can be tricky.

As against the CAG’s calculations on presumptive losses, the UPA had made three criticisms – that there was “zero-loss” as government had the right to set goals other than revenue maximisation in allocating spectrum; moreover, CAG was wrong to use the prices obtained in a 3G auction in 2010 to calculate spectrum losses in 2008; and, in any case, the failure of the November 2012 auctions showed that setting high reserve prices was not the answer. Sibal and I&B Minister Manish Tewari taunted CAG after the November flop: “Mr CAG, where is the Rs 1,76,000 crore?”

However, CAG’s last missive under Rai—sent to the Department of Telecom (DoT) on 20 May, two days before his retirement—makes two points that make it abundantly clear that the spectrum losses under Raja were real and not just notional, and that there must have been a scam.

As reported by Firstpost yesterday, the CAG’s letter alleges that last November’s spectrum auctions were a flop due to possible cartelisation (which is circumstantially provable by many operators’ failure to participate despite wanting more spectrum), but the more telling observation is how the spectrum allocated by Raja at throwaway prices has just been hoarded. If cheap spectrum was hoarded or left unused, that leads to a further loss to the exchequer, since lack of usage means government loses out of spectrum usage charges.

So CAG has come up with yet another loss figure: Rs85,014 crore loss due to cartelisation and inefficient use of spectrum by many of the Raja-allotted telcos. In the letter to DoT, it said that the “inefficient allocation and inept management of 2G spectrum” and the cartelisation that led to flops at the last two auctions meant the government was unable to “fruitfully utilise 453.50 Mhz of second-generation airwaves for the last one year,” reportsThe Economic Times.

So to the charge of zero-loss, CAG is effectively saying, “No sir, even at the prices at which Raja sold spectrum there is a loss on revenue share due to under-utilisation of allotted spectrum.”

CAG’s reasoning: operators who have been clamouring for scarce spectrum for long didn’t bid for spectrum when it was offered offered; on the other hand, operators who got cheap spectrum from Raja in 2008 were underutilising it. Reading between the lines, one can deduce that the unstated purpose of selling spectrum at prices determined in 2001 was to create rent-seeking opportunities for cronies who never intended to use the spectrum efficiently in a hurry.

As Firstpost wrote yesterday, to support its stand on cartelisation CAG pointed out that players clamouring for spectrum didn’t bid at all in the November auction. Bharti made 68 applications for additional spectrum for 21 service areas (excluding Delhi) between 2008-2011 but bid only for 1.25 Mhz in Assam in the November auction. It had also sought additional spectrum in Delhi, but did not even bid for them. In fact, Sunil Mittal, Bharti chief, made the prediction that the November auctions would be over in one day. How did he know?

Ditto for Vodafone, which sought extra spectrum in Delhi through 15 applications over nine years starting 2003, but got none. It didn’t bid for Delhi last November. The auctions were a flop only because there were no bids for the big circles such as Mumbai and Delhi, which is exactly where spectrum is scarce.

On the other hand, look who got Raja’s spectrum and didn’t use enough of it: Uninor, Videocon, Loop, Sistema Shyam and Etisalat got the bulk of the licences issued in 2008 (105 out of 121), but they added only 10 percent of the additional users in the four years from January 2008 to 2012.

CAG also pointed out that Bharti was servicing 90 lakh subscribers in Delhi with 10 Mhz, while two other service providers in the same area had 12 Mhz and 4Mhz each – but servicing just 26 lakh subscribers and 28.51 lakh.

CAG said: “It was therefore no wonder that a top TSP (telecom service provider) which had not participated in auctions held in November 2012 and March 2013 opted to have a private tie-up with two TSPs for sharing 2G spectrum in many service areas in 2013 rather than approaching DoT for award of spectrum through auction.”

It is also clear where Bharti got some of the additional spectrum it needed: from the “overpriced” 3G auctions in 2010.

The headline story on 3G so far has been that it has failed. Kapil Sibal said as much last yearwhen he inaugurated  Bharti’s 4G services. “3G has not delivered because they (the bidders) paid such huge prices for the spectrum and there is no liquidity in the market for them to invest in the infrastructure and the devices to deliver 3G, for which 2G was successful and 3G was not successful.” Sibal said this largely to support his theory that 3G was overpriced, and hence the 2G scam under Raja cannot be calculated from the 3G prices.

But the truth is more nuanced.

A report by Credit Suisse Securities says: “Contrary to a widely-held belief that predatory pricing in 2G data is killing the 3G data opportunity, recent numbers from Idea show that the realised 2G data tariffs are actually at a 56 percent premium to 3G realisations. We believe this is because of smart 2G data packaging. In fact, this pricing strategy could encourage customers to upgrade to higher data plans (even 3G) once they get hooked on 2G data – a trend clearly visible in the increasing data ARPU (average revenues per user) trend of the past few quarters.”

What does this imply?

First, if 2G is generating higher revenues than 3G, comparing 2G spectrum pricing to 3G by the CAG was not that far-fetched.

Second, if Idea can earn more from 2G than 3G, it means the 3G potential is under-exploited. In fact, Bharti is said to be using the high-cost 3G it won in 2010 to shift voice calls—which is probably why it didn’t bid much in 2012 and in March this year. It is optimising the spectrum already owned. This may go against CAG’s cartel theory, but it supports the fact that spectrum is scarce—and that’s a reason for not gifting it away like Raja did.

Third, concerns over overpricing of the reserve price in the recent auctions may thus be overblown. Telecom is not meant to be an instant profit generator. Operators have to build demand and, over the medium term, high reserve prices will be covered by smart pricing of products – as the Idea experience shows.

Says the Credit Suisse report: “In our analysis, most of the top cities in 3G circles are already under 3G coverage for leading telcos, and 13-14 percent of mobile voice traffic is already flowing through 3G networks. This 3G voice offload is helping free up precious capacity on the spectrum-starved 2G networks – opening up bigger opportunities, particularly on 2G data.”

3G may be generating only 3-4 percent of revenues, but it is being used, and helping meet spectrum demand.

From all three angles—zero-loss, wrong comparison on 3G by CAG, and the failure of the recent spectrum auctions due to high reserve prices (widely blamed on CAG)—Sibal may be partially or fully wrong. There is indeed a huge need for spectrum, and telecom packages marketed well seem to be able to generate the right revenues—whether 2G or 3G.

The auction failure may or may not be the result of cartelisation, but it may only need the return of better business optimism and sharper marketing on 2G and 3G products to make the current reserve prices profitable over the medium term.

Sorry, Sibal, your observations do not hold water.