With the government ready to auction 3G and BWA spectrum, Sunil Jain met Qualcomm’s India and South Asia President Kanwalinder Singh to talk about whether the budget target of Rs 35,000 crore from the auction was realistic; did a low-base price for Broadband Wireless Access (BWA) make sense, could standalone players survive and so on. Qualcomm is the world’s biggest technology provider in the 3G space. Excerpts:
Are you happy with the progress of 3G — BSNL/MTNL haven’t got too many customers, but the Tata/Reliance EVDO internet cards are doing well.
BSNL/MTNL are currently focused on rolling out their 3G networks and are seeing a good initial response. Tata/Reliance are in 40-plus cities already and they’re adding more every day. The response to EVDO internet dongles has been good and now Reliance has even started offering pre-paid internet dongles so you can control how much downloading you want to do. We’re getting the price of dongles to where we want them — we are heading towards the sweet spot of Rs 2,500 in the market. We expect to do around one million-plus volume in this calendar year and grow from there.
Will the government get the Rs 35,000 crore it hopes for from the 3G/BWA auctions?
The voice revenue per minute in India is among the lowest in the world at around 50 paise. Though voice provides the necessary foundation for existing players, the revenue growth lies in broadband. While the government should set a reserve price for 3G/BWA that is fair and ensures a quick roll-out, let’s recognise that both the spectrum acquisition cost as well as capital expenditure are small factors in an incumbent’s 3G business case since they will be amortised over 20 years or so. The government has a total of five 3G slots and 4 BWA slots — assuming the government finally decides on a BWA base price that’s a fourth of 3G, this means there are six 3G slots (four of BWA equals one of 3G) being auctioned. So that’s Rs 6,000 crore apiece over 20 years.
The 3G capex will differ depending on whether an existing player is coming in from the CDMA or GSM side, but the maximum an existing player needs to invest is around $10,000-12,000 per base station — since incumbent players need to deploy electronics on their existing towers, their only costs relate to the electronics. So, let’s take Bharti which has the highest number of base stations — around 100,000 — at $12,000 apiece, that’s an expenditure of $1.2 billion (Rs 6,000 crore) at an all-India level.
Is there a business case for new 3G/BWA entrants?
Not at all since the new entrant will either have to set up new towers or hire them — either way, it cannot compete with the incumbent players. It gets worse for someone doing Wimax since this does not integrate with any 3G or 2G technologies — it does not do either voice or broadband well in a mobility context. And since it is being deployed at a higher frequency than what the existing 2G networks use, it will require additional base stations (compared to 3G) for full coverage in an area. At best, Wimax may end up serving fixed urban broadband niches, where it will soon be challenged by LTE, which has worldwide momentum and integrates seamlessly with 3G and 2G, and therefore will not need so many base stations. Ironically, the whole Wimax pitch has been centered around rural India and affordable broadband where it cannot match what 3G mobile can offer.
What about the operating expenditure?
Rentals are around Rs 30,000 per month per tower site today and you can add around another Rs 25,000 for diesel and other pass-through costs. Let’s assume a Wimax player in a city like Pune deploys the same 550 towers that an incumbent operator does. Let us also assume that each base station yields 100 broadband customers. Both these assumptions are extremely generous towards a Wimax firm. With these assumptions, the site cost per subscriber works out to Rs 550. Add the other operational costs of running a business and you would find that even if the Wimax firm gets Rs 750 per subscriber as revenue, the business case is negative just on the opex. For incumbent players (with 3G/LTE), the opex is already paid for by voice subscribers, so the broadband revenue is the upside.
So there is no case for standalone broadband for rural areas? But the reserve price for BWA has been kept much lower than that of 3G on the grounds that this will encourage broadband activities.
In the final analysis, lower spectrum reserve price for Wimax is irrelevant since a Wimax network cannot cover opex outside dense urban pockets. Affordable broadband in rural India can only be provided by 3G technologies where voice customers help defray the costs of providing services.
More From This Section
One theory is the lower reserve price will ensure BWA spectrum gets bought cheap. Then, when full-fledged internet telephony is allowed, companies will offer normal voice services on this cheaper BWA spectrum as well.
You’re right in the sense that packet voice, which is commonly referred to as Voice over Internet Protocol, can technically be supported on 3G/BWA. But the packet voice business case is unviable due to the additional investments that will be required at each telco’s switching centers.
What matters is the technology you deploy to offer both voice and high-density data capabilities, and whether there is a business case for it. If you do Wimax for data, it cannot provide voice services unless large investments are made in additional base stations. To go back to the Pune example, the telco will need to deploy Wimax on every base station in the area and it will need extra base stations as well — if it doesn’t do this, the Wimax device will not work. Apart from the additional cost, it is not likely that there will be enough data-customers in each base station area. Instead, if it deploys LTE in only those base station areas which require high data capacity, the same LTE phone can be used for voice and lower data-speed applications in other base station areas where 3G has been deployed.
Should the government just have the same base price for BWA as it has for 3G spectrum — can 3G devices like a Blackberry Bold which works on 2100 MHz, work on the 2300 or 2500 MHz BWA spectrum?
LTE, specifically the Time Division version, can be deployed in the BWA spectrum being auctioned in India. LTE devices — and I am sure there will be an LTE Blackberry or iPhone in the future — will be able to enjoy ultra broadband experience in dense urban pockets, while seamlessly integrating with 3G broadband in the rest of the cities and even with 2G data in a remote village. This kind of voice and broadband service experience cannot be provided by other technologies. So, yes, LTE will eventually end up using the BWA spectrum optimally in India.
Through all of this, we’ve assumed 3G handsets will be available at affordable prices.
We can get an ex-factory price of Rs 5,000 for a 3G handset which makes it competitive and mass-market. So we’re there at the sweet spot and the addition of data offers service providers the opportunity to increase their ARPUs significantly.