India - Telecom

09 July 2012

Regulatory update

Institutional Equities
Figure1: Snapshotofspectrumpayoutsbytelcosundertheproposedoptions TotalPayout(Rsbn) Option1 Option2DD MMM YYYY Option3* >4.4MHz (GSM) >6.2MHz(GSM) >0MHz />2.5MHz(CDMA) />5MHz(CDMA) Bharti 111 57 32 Vodafone 88 39 19 Idea 105 28 12 Dualtechnologylicensees 387 126** 115 BSNL/MTNL/Others 368 141 83 Industrytotal 1,059 392 262
Source:TRAI,DoT,Mediareports,IIFLResearch;CalculationbasedonTRAIspricesand800MHzand 900MHzareat2x1800MHz;*Inoption3,weassumethatTataandRCOMpayfortheirentireGSM spectrum holdings; **In option 2, if RCOM and Tata surrender excess spectrum in CDMA above 2.5MHz,theywouldrequiretopayoutonlyRs3.6bn

More unworkable options from DoT

As per the media, DoT has proposed three excess spectrum payment options payment for all spectrum for remaining licence life, payment for excess spectrum for GSM/CDMA over 4.4/2.5MHz and payment over 6.2/5MHz (but not both GSM and CDMA). As we explain in our note, none of these are feasible. As a result, we expect the government to become increasingly defensive and incapable of enforcing punitive regulation (high payouts only for GSM incumbents, and re-farming). Decision making around auctions and excess spectrum seems set for more delays, and this should act against cancelled licensees. Three options mulled for excess spectrum charges: The three options are charging for 1) all spectrum prospectively till expiry, 2) spectrum beyond 4.4/2.5MHz for GSM/CDMA and 3) spectrum beyond 6.2/5MHz for GSM/CDMA. Our calculations based on TRAI prices (auction prices should be 30% lower) validate the media reports. Additionally, we discover Bharti would pay Rs111/57/32bn under options 1/2/3. The corresponding numbers for Idea are Rs105/28/12bn. None of the three options workable: RCOM and Tata would need to pay much higher under option 1 and hence this has already run into political opposition. Options 2 results zero payouts for dual-technology companies if they surrender CDMA spectrum above cutoff, which they can easily manage, in our opinion, and hence this is potentially equally unacceptable, unfeasible. In option 3, media reports seem to imply a twist if one pays for GSM, CDMA spectrum becomes free (which again favours dual-technology companies). Anything free will certainly attract controversy, especially if it is the CDMA spectrum which until now was being priced at 2x 1800 GSM. Hence all options seem unworkable. Complete reset, lower reserve prices the best option in our view: Full immediate payment for 20 years would be a more feasible option. This would ensure a level-playing field, eliminate scam allegations and would be revenue generating for the government. At our assumed price, industry would see $42bn and deferred payment/ spectrum trading / mortgaging could help with financing. This clean solution is however dependent on successful auctions, and that needs lower reserve prices. We are beginning to believe that this is inevitable.
G V Giri | gvgiri@iiflcap.com 91 22 4646 4676

Figure2: EndFY13debt/FY13EbitdaRCOMatmaximumstretchunderalloptions

FY13netdebt/Ebitda Current 12.0 10.0 8.0 6.0 4.0 2.0 0.0 Bharti
Source:DoT,TRAI,IIFLResearch

Option1

Option2

Option3 10.0 7.6 7.5 6.3

2.1 2.5 2.3 2.2

3.6 2.0 2.4 2.2

Idea

RCOM

Balaji Subramanian | balaji.subramanian@iiflcap.com 91 22 4646 4644

Institutional Equities

India - Telecom Figure3: Option1wouldburdenallplayers,especiallydualtechnologyplayerssuchas RCOMandTatawhileoptions2and3wouldfavourtheseplayers Operator TotalPayout(Rsbn) Option1 Option2 Option3* GSM >0MHz >4.4MHz >6.2MHz Bharti 111 57 32 Vodafone 88 39 19 Idea 105 28 12 Aircel 109 4 2 RCOM 72 4 1 Tata 43 0 0 BSNL/MTNL 192 130 81 Total 721 261 147 CDMA >0MHz >2.5MHz** >5MHz RCOM 147 72 0 Tata 124 51 0 BSNL/MTNL 67 8 0 Total 338 131 0 Total Bharti 111 57 32 Vodafone 88 39 19 Idea 105 28 12 Aircel 109 4 2 RCOM 220 75 72 Tata 167 51 43 BSNL/MTNL 259 137 81 Total 1,059 392 262
Source:TRAI,DoT,Mediareports,IIFLResearch;CalculationbasedonTRAIprices;*Inoption3,we assume that Tata and RCOM are required to pay for their entire GSM spectrum holdings; **In option2,IfRCOMandTatasurrenderexcessspectrumabove2.5MHzinCDMA,theywouldneedto shelloutonlyRs3.6bn

Three unworkable choices: Media reports suggest that DoT has proposed three options for the basis of calculation of excess spectrum charges, using TRAI prices provisionally. The matter is with the EGoM. The options are: 1. All telcos pay prospectively (from now till licence expiry) for all spectrum held by them. 2. Telcos pay prospectively for GSM spectrum above 4.4MHz and for CDMA spectrum above 2.5MHz. 3. Telcos pay prospectively for GSM spectrum above 6.2MHz OR for CDMA spectrum above 5MHz (our interpretation based on media reported numbers). While the earlier round of discussions also included an option of telcos paying for all spectrum for the next 20 years so that all licences across circles become co-terminus, recent media reports do not suggest that this is being actively pursued.

We calculate excess spectrum charges based on TRAIs 1800Mz

reserve prices of Rs36bn/MHz and 800/900MHz at 2x of 1800. We assume that these are effective from October 1, 2012. show that under option 1, the government would make Rs721bn on GSM spectrum and Rs338bn on CDMA spectrum.

Our calculations, which agree with media reported numbers,

We also confirm the media numbers around option 2

GSM/CDMA payouts would be Rs261/Rs131bn. However, we think that if they surrender excess CDMA spectrum above 2.5MHz, they would practically pay nothing. We interpret that the government is likely to ask dual technology licensees to pay for all spectrum in one out of GSM/CDMA while the other would be free. Our calculations based on this imply Rs262bn GSM industry payment, largely in-line with media.

Media numbers around option 3 show zero payments for CDMA.

gvgiri@iif lcap. com

Institutional Equities

India - Telecom

Extreme industry leverage makes these options fanciful

Figure4: Industry ex top three (Bharti, Idea, Vodafone) has Rs1trn debt and Rs2.8bn Ebitda Company(Rsm)asofFY12 Revenue 3yrrevcagr Netdebt EBITDA Debt/EBITDA Bharti 714,508 24.6% 650,998 236,573 2.8 Idea 195,411 24.1% 130,875 50,924 2.6 VodafoneIndia 332,122 18.0% 357,000 84,262 4.2 RCOM 193,380 4.6% 362,493 54,461 6.7 Aircel* 65,312 26.5% 170,000 20,000 NA Uninor 28,631 NM 80,000 19,834 NA Tata* 119,062 12.6% 150,000 8,168 NA Sistema 14,035 NM 57,746 17,309 NA GTLI 13,933 85.0% 117,144 7,530 15.6 Viom* 13,500 NA 80,000 6,075 13.2 TotalA 5.8 1,689,893 2,156,256 374,515 Top3B 1,242,041 1,138,872 371,759 3.1 Others=AB 447,852 1,017,383 2,755 369
Source: Company, TRAI, Industry sources, IIFL Research; *Ebitda figures for Tata, Viom and Aircel areestimated;RevenuefiguresforTataandAircelbasedonTRAIAGRreport

Option 2 allows dual-technology companies to avoid any payment a problem in an environment of bloodlust Under option 2, RCOM and Tata could avoid large CDMA payments (US$1.2bn and US$900m) by returning excess over 2.5MHz of their CDMA spectrum. This is possible since they could easily carry on dongles business for high speed data on 2.5MHz and transfer voice business over to GSM a migration of 50m CDMA voice subscribers with full subsidy on GSM handsets of $5 would cost merely $250m, an easy choice in our view. For GSM, their holdings would largely be under cut-off, so option 2 would entail almost no spectrum outgo for dual-technology companies. This seems politically unfeasible given that free spectrum is the genesis of the 2G scam. Option 3 involves substantial free spectrum to some extent very brave in a scam charged environment Option 3 seems to give dual technology operators the choice of keeping their entire CDMA spectrum by paying for their entire GSM spectrum. This also is favourable to dual technology licensees as they get to keep both GSM and CDMA spectrum while they pay for only one of them. Since BSNL and MTNL would need to pay but cannot, we presume they will surrender spectrum, which means that a good chunk of spectrum will come into the market over time. This option will also entail fairly low returns to the government excluding BSNL / MTNL, assuming 30% less than TRAI prices, the government may net less than Rs150bn. This plan too looks weak as there may be no takers within the government. Lower base price and simultaneous reset the best option in our view According to an earlier DoT proposal, all operators must pay an auction determined price for the next 20 years, regardless of the remaining years in the current licence. While media reports indicate that this is not being actively explored, this is the best way forward in our view. All licences will be reset and become co-terminus
3

Option 1 would burden everyone, especially dual technology players hence a non starter If the government decides to go with option 1 (all spectrum charged till expiry), the incumbents (Bharti, Vodafone and Idea) stand to benefit the most as many of their licences are only 3-4 years away from expiry. RCOM and Tata would face a spectrum outgo bill around US$3-4bn as their licences are relatively new. Given their high leverage, it looks unlikely that they will able to make these payments and they would be faced with the unpleasant choice of entirely surrendering one of their spectrum holdings (either 4.4MHz GSM in 1800 band or 2.5-5MHz CDMA in 800. This plan has already encountered political opposition. We consider it a non-starter for the above reasons.

gvgiri@iif lcap. com

Institutional Equities

India - Telecom

ensuring level-playing field in unambiguous terms, and leave little room for complaints by anyone. This proposal would substantially eliminate regulatory controversy, be viewed as revenue generating for the government. This would also liberalise all spectrum allowing telcos to deploy any technology of their choice in the future. This would also obviate the need for refarming as 900MHz spectrum would have been paid at market price.
Figure5: Full 20 yr payout using TRAI prices. While all operators paying for spectrum afreshfor20yearswouldbethebestoptioninourview,lowerbasepricesanddeferred paymentwouldberequiredtomakethisfeasible Telco Rsbn Bharti 454 Vodafone 484 Idea 333 Aircel 202 RCOMGSM 176 RCOMCDMA 348 RCOMTotal 524 TataGSM 113 TataCDMA 319 TataTotal 432 BSNL/MTNLGSM 604 BSNL/MTNLCDMA 208 BSNL/MTNLTotal 812 Industrytotal 3,242
Source:DoT,TRAI,Mediareports,IIFLResearch;BasedonTRAIsprices

prices fearing future scam charges. The government is yet to find an auctioneer and has pushed out auctioneer RFP response to 20th July. Hence we do not expect these decisions to be made quickly and consequently do not see auctions happening in FY13. A delay in auctions is negative for cancelled licensees. Under the current circumstances, cancelled licensees will be hard pressed to continue capex, and that should make for better market share movement for Bharti and Idea.
Figure6: Sectorvaluationmatrix Item CMP(Rs) TP(Rs) Rating P/E(FY13) P/E(FY14) P/E(FY15) EV/Ebitda(FY13) EV/Ebitda(FY14) EV/Ebitda(FY15) Ebitda/Int(FY13) Netdebt/1yrfwdEbitda Netdebt/1yrfwdFCF RevenueCagr(FY12+3) EbitdaCagr(FY12+3) EpsCagr(FY12+3)

Bharti 321 364 BUY 19.6 13.0 10.8 7.0 5.6 4.6 6.9 2.2 9.8 14.3% 17.7% 38.6%

Idea 85 114 BUY 20.0 13.5 9.7 5.6 4.6 3.3 5.5 1.8 NM 19.0% 24.4% 59.0%

RCOM 69 54 REDUCE 38.0 30.5 15.2 8.0 7.7 7.2 2.6 6.0 11.2 5.1% 7.1% 0.5%

Source:IIFLResearch;PricedasatcloseofJuly6,2012

While the above $60bn figure seems intimidating, at 30%-40% lower than TRAI prices (which we expect spectrum auctions to throw up), plus deferred payment/spectrum trading and mortgaging, the payouts might be manageable, especially in a moderately firmpricing environment. But will auctions happen at all in FY13? The biggest obstacle however is the reluctance of the government to fix lower reserve
gvgiri@iif lcap. com 4

Institutional Equities

India - Telecom

Key to our recommendation structure BUY - Absolute - Stock expected to give a positive return of over 20% over a 1-year horizon. SELL - Absolute - Stock expected to fall by more than 10% over a 1-year horizon. In addition, Add and Reduce recommendations are based on expected returns relative to a hurdle rate. Investment horizon for Add and Reduce recommendations is up to a year. We assume the current hurdle rate at 10%, this being the average return on a debt instrument available for investment. Add - Stock expected to give a return of 0-10% over the hurdle rate, ie a positive return of 10%+. Reduce - Stock expected to return less than the hurdle rate, ie return of less than 10%.

Published in 2012. India Infoline Ltd 2012. This report is published by IIFLs Institutional Equities Research desk. IIFL has other business units with independent research teams separated by Chinese walls, and therefore may, at times, have different or contrary views on stocks and markets. This report is for the personal information of the authorised recipient and is not for public distribution. This should not be reproduced or redistributed to any other person or in any form. This report is for the general information of the clients of IIFL, a division of India Infoline, and should not be construed as an offer or solicitation of an offer to buy/sell any securities. We have exercised due diligence in checking the correctness and authenticity of the information contained herein, so far as it relates to current and historical information, but do not guarantee its accuracy or completeness. The opinions expressed are our current opinions as of the date appearing in the material and may be subject to change from time to time without notice. India Infoline or any persons connected with it do not accept any liability arising from the use of this document. The recipients of this material should rely on their own judgment and take their own professional advice before acting on this information. India Infoline or any of its connected persons including its directors or subsidiaries or associates or employees shall not be in any way responsible for any loss or damage that may arise to any person from any inadvertent error in the information contained, views and opinions expressed in this publication. India Infoline and/or its affiliate companies may deal in the securities mentioned herein as a broker or for any other transaction as a Market Maker, Investment Advisor, etc. to the issuer company or its connected persons. India Infoline generally prohibits its analysts from having financial interest in the securities of any of the companies that the analysts cover. In addition, the company prohibits its employees from conducting F&O transactions or holding any shares for a period of less than 30 days.

gvgiri@iif lcap. com