How The Mighty Have Fallen

When 3G auctions were over on May 19, there was little fanfare to herald the new era in telecom, that of super-fast video downloads or electronic gaming possibilities. The mood across most operator camps — Bharti, Vodafone, Idea and even Reliance — was sombre rather than jubilant. Bharti even put out a sullen press release blaming everything from the auction format, spectrum shortage to policy uncertainty for its inability to win pan-India 3G spectrum. Vodafone’s head of strategy, Samaresh Parida, says, “Due to artificially constrained supply, the 3G auction prices were driven much higher than
expected. While the government garnered huge amounts from the auctions and the sector shares substantial revenues with it, I fear that the industry’s future could be in trouble because of high spectrum prices, large number of competitors and lack of incentives to consolidate.”

“My personal view is that the next 24-36 months will be most difficult for telecom companies but good for subscribers,” says Romal Shetty, head, telecom practice, KPMG. Super if you are planning on buying a new SIM card, not if you wanted to buy any of the telecom stocks. Most fund managers are now beating a hasty retreat from the sector. “I have zero telecom stocks in my portfolio,” says Rajiv Thakkar, CEO & director, Parag Parikh Financial Advisory Services.

So why is a business that is so great for the consumers, a bad business now? Most telecom companies will now be adding significant amounts of debt to their balance sheet. For every rupee of equity, Bharti only had 75 paise of debt. It will now have almost Rs. 3 of debt for a rupee of equity. Yes, Bharti is different in that much of its debt is because of purchase of Zain’s Africa assets a few months ago. But look at Idea. Its debt to equity ratio will go from 0.67 to 0.96 in a year’s time. Reliance Communications’ balance sheet is already stressed and will not improve after it pays the 3G bill.

All this increase in leverage comes at a point where the money that they make has been steadily declining. Indian telecom operators say that they are in the business of manufacturing “minutes”. Well, as more and more operators have been granted licenses, the volume of “minutes” has gone up steadily.

In 2008, the industry made 250 billion minutes. Currently, it makes 600 billion. In 2008, it sold each minute for 60 paise. Today, it sells it for about 35 paise. The running cost of making minutes — without taking sunk costs into account — is now stuck at around 25 paise a minute. And this is for Bharti, which has the lowest cost. For others it is even higher. It is not surprising then that there is going to be much free cash flowing out of the balance sheets of most telecom companies. Bharti, the one exception till now, will also soon cease to remain one.

One could argue that these things are merely a blip, a part of a cycle, and that the companies will eventually find a way to make money. But that’s just optimism. The reality is that the old order of simply making minutes and selling them to ever willing customers is over. Telecom operators will have to learn new tricks and that could take some doing.

How the Body Blows Came
The reason why it will be more difficult is because almost all the parameters — customers, competitiveness, regulation and raw material prices — that define this industry have turned adverse. The humungous multitude of customers was one huge pain reliever for this industry. But they don’t make customers like they used to. You can’t blame them. After all, if the business of telecom companies is making minutes then the customers are only “minutes” shopping. “We think almost 30 percent of customers today use more than one SIM card,” says a senior executive at a top-4 telecom firm.

Sandeep Barasia, a partner with consulting firm Bain says he knows people who own 17-18 active SIM cards. This means customers keep one primary SIM card that gives the constancy of purpose and use other cards to make cheap phone calls depending upon the tariff plan of the secondary SIM provider.
That shows up in a metric called Average Revenue per User (ARPU). For instance, Airtel with an ARPU of Rs. 220 will mostly be a primary SIM and Uninor at an ARPU of Rs 92 the secondary, tertiary or even quaternary SIM.

This is only to be expected given that most circles now have as many as 12 competitors. Many of them have come into the market over the last five years. “How do you differentiate if voice or ‘minutes’ are all that you sell,” says the telecom practice head of an investment bank.

The answer is simple. You cut price either directly or disguise it through a clever tariff plan and hope that you sign on subscribers to build a critical mass. That at least gets you into the game. This is exactly what Tata DoCoMo did with its per second billing in October last year. Bharti responded in kind by cutting tariff on various plans. Soon everybody jumped into the mud-bath. The result was not pretty. Telecom earnings dropped sharply. All the top operators’ average revenue per operator dropped down anywhere between 7-10 percent in each quarter for the last four quarters. And all of them saw a continuous decline of 6-7 percent in each of the last quarters in revenue per minute that they sold.

That’s putting pressure on the financials because the cost of manufacturing that minute is not declining too much. “Earlier the profit per minute was never ever calculated. Now companies are looking at that metric as well, otherwise the numbers don’t make sense. Managers who were judged only on subscriber addition earlier now have to deliver profit per minute as their key parameter. Many people have been fired in the bigger companies on that account,” says Shetty of KPMG.

The bad news is that the cost of making that minute will go up in the short term. The key reason is the high prices that all operators have paid — 9 times the base price in Mumbai and Delhi and 5.7 times in Bihar — in the recent 3G auctions. No one expected such high prices — not the government, not the scores of telecom consultants, not the analysts or fund managers, not even the operators themselves. “Everybody was going with 2-2.5 times the base price. What’s now happened is, because of competition they’ve gone beyond even stretch targets. It was about ‘If I don’t have it today it might become costlier and I may lose out later,” says Shetty.

While the data revolution will happen, for the moment, all operators have used voice revenues to bid for the auctions. Over the last seven years this is the biggest chunk of spectrum [electromagnetic frequency on which mobile companies transmit signals] that the government has released. All the governments have been unable to make the right amount spectrum available in a consistent fashion. Naturally, then prices paid in 3G auctions have a scarcity premium because most of the 12 operators wanted it. “Till 2012-13, the serious players/biggies are clear that subscriber acquisition is the only thing. They wanted 3G primarily for spectrum [voice],” says Shetty.

Spectrum costs aside, there will be fresh capital expenditure to upgrade their infrastructure to support 3G technologies, between $500 million to $1.5 billion depending upon the vintage of the player, and since the customers will not agree for an increase in price on voice, expect profitability to decline. Analyst estimates suggest that over the next 18 months, EBITDA (earnings before interest, taxes, depreciation, and amortisation) for most companies should decline another 2-3 percent from where they are at present. Not that the present was a particularly great place to be in.

For instance, Reliance has seen its EBITDA slide from 41 percent to 31 percent in just the last six quarters. Given that debt on the books is only going to go up, old telecom hands believe that the sector will trot up losses on a net basis, except for one or two top players. “A lot of new players are asking themselves if they can earn their investment back,” says a former CEO of a telecom company.

Déjà Vu
The interesting thing is that this industry has been through a similar period of distress just about a decade ago. At that point no one was making money (collective losses of $1 billion, pretty large for the time) and it looked like mobile telephony would remain a rich man’s game. In hindsight, it was government policy and regulation that was responsible.

Between 1995 and 1997, the government handed out over 40 separate licences to operators to roll out GSM networks across four metro cities and 18 states. The new operators had to pay a flat license fee to the Indian government for three years regardless of the money they earned from subscribers, which too was limited because the tariffs were dictated by the Department of Telecom (DoT).

Worse, most of them in their eagerness to capture what they thought was a large, lucrative market, overbid massively, committing more than Rs. 27,000 crore over 15 years. In a couple of years, many were already defaulting on these commitments because the market didn’t take off the way they expected, primarily due to high prices for handsets, telecom equipment and calls.

In was only with the formulation of the National Telecom Policy of 1999, NTP-99, that telecom players discovered that elusive financial metric called “profit”.

NTP-99 did away with a fixed license fee in favour of charging operators roughly 12-13 percent of their annual revenues. This was brought down to 8-10 percent in 2001. While bringing in newer operators — adding up to four in each circle — the government gave each one more freedom to set their own tariffs.

But what the government gave away it got back in return. Since the existing operators had not been able to keep their bargain and pay upfront, they had to agree to more competition being let loose in the industry. As a result, India now has 12 and sometimes 15 competitors in one circle. Now clearly competition is good. It keeps the companies sharp and benefits consumers.

However, all aspects of the market have to be working at optimal levels for competition to have desired effects. In India, two aspects: Supply of raw material that makes “minutes”, namely spectrum is severely constrained. And two, it is very difficult for people get out of business. As a result, what we have today is an industry that usually thrives on scale having the feel of a small scale sector.

War of the Policymakers
And the root of this is perhaps the difference in opinion between the way two arms of the government, DoT and TRAI, that regulate telecom think. TRAI was created in 1997 to prevent a conflict of interest in the case of DoT — both a regulator as well as a competitor (the state-owned telephony provider, before BSNL and VSNL were carved out of it).

DoT initially tried to have TRAI as one of its own divisions, failing which it stalled TRAI’s very first order on call tariffs. It did this by invoking an arcane clause in the TRAI Act that allowed the government — DoT — to override TRAI on matters related to policy. The two have been at loggerheads ever since, with only the intensity varying.

NTP-99 sought to correct this regulatory clash by clarifying the ambit of TRAI’s authority in a clearer manner. But it also added to the confusion by awarding TRAI’s power to adjudicate disputes to a newly created body, the TDSAT. So operators now had the TRAI for regulation, the TDSAT for legally disputing
what they didn’t like and the DoT framing government policy.

Cut to the present and TRAI’s recent recommendations on 2G spectrum pricing and consolidation have been met with overwhelmingly critical responses by all and sundry in telecom. “For the first time I’m seeing TRAI regulations that are meaningless. It’s almost as if a smaller player said that M&A norms should be anti large players or almost as if they are targeted at big GSM players like Bharti and Vodafone,” says Sridhar Pai, the founder of telecom consulting company, Tonse Telecom.

The other stunning recommendation from TRAI is a sneaky one. It recommends that any further allocation of frequency from the old stock of spectrum [2G] be made at the prices at which companies have won their share in the new stock [3G] spectrum.

“Yes, the incumbent operators took it free of cost and enjoyed it; they should have paid for it. But asking them to pay 3G market determined price in 2010 is a little unfair. The 3G auctions story would have been different if companies were told beforehand that 2G prices would be linked to 3G prices,” says Shetty.

Not surprisingly, operators have chosen to question TRAI’s recommendations. Airtel and Vodafone both wrote to the Communications minister A. Raja, the former calling the recommendations “shocking, arbitrary and retrograde”. The two, along with Idea, then challenged the recommendations before the TDSAT, which told them the difference between recommendations and orders.

Radio spectrum is a scarce resource all around the world, in India it is much scarcer simply because there are more operators among whom it is divided. But if TRAI’s goal in slicing ever smaller pieces of the spectrum pie to more players was to increase competition, the net result might be the exact opposite.

According to a study commissioned by the GSM Association in December 2008, the average Indian operator with 5.5 MHz spectrum has one quarter the 22 MHz spectrum available to operators around the world.

 

That means successful operators like Airtel or Vodafone reach the spectrum limits the fastest and then suffer the consequences of network congestion: Losing customers to smaller competitors. That isn’t better competition; it’s government-enforced socialism of market share. Better competition would be if those customers left them for lower prices or better services. No wonder then that operators like Airtel and Vodafone plan to deploy a significant majority of the 3G spectrum they bought to provide “defensive voice relief”, meaning use 3G for voice calls so your best customers don’t leave you for better voice quality.

Older players have also been hit by a slew of other TRAI recommendations — licence renewal fee, 2G spectrum charges, 900 MHz re-farming, cap on consolidation — and they are not allowed to share spectrum. Newer players come in with a reduction in licence fee, a government-given market value for their spectrum (never mind subscribers), spectrum sharing has been allowed for them.

“Net effect of TRAI recommendations is that incumbents have been put at greater risk: Lesser spectrum, 900 MHz advantage being taken away, lesser M&A options,” says Mahesh Uppal, a telecom policy consultant.

Now why should it be like that?
The only reason is that Indian regulators are fascinated with the idea of “large” and “vibrant” markets and a desire to constantly level the playing field to make sure that new operators feel at home. But if there is freedom to enter the industry there must be freedom to exit it as well. And that’s why the M&A recommendations that TRAI has suggested make so little economic sense.

Consider some proposals. A merged entity cannot have more than 30 percent market share nationally. The total spectrum held by a merged entity cannot exceed a certain point, which is not very high. “How can you decide on 30 percent? It is just not comprehensible. And if I have to return spectrum then why should I merge?” asks a telecom industry veteran.

In most telecom markets across the world, generally there are four to five players. The top two make money, the third makes just enough to keep body and soul together and the number four and five sell out getting newer guys into the game who think they can rock the boat. Since in India there are very few incentives to merge, the number of competitors will not come down by a lot.

 

 Infographic: Malay Karmakar

“If policy enables, we should get down to 6-7 players in two-three years — the top four big guys and younger/more agile competitors (with foreign funding) and BSNL/MTNL. I see a decent role for six-seven operators but not 15,” says Kunal Bajaj of Analysys Mason, another telecom consulting firm.
Sachin Gupta, a financial analyst with Nomura agrees and says you need to burn serious to find out whether you can. “Very few can exist profitably, perhaps five-six. More than consolidation there might be elimination, forcing some to pack up and go. Smaller players must spend $0.5-1 billion to get more clarity to see if they can survive and make money,” he says.

If you can’t exit then companies will stay and compete destructively. TRAI recommendations have given a fresh breath of life to the newer entrants. They can now stay on longer before deciding whether to quit or not. “If you are a Datacom or Etisalat you will like these recommendations. You are compounding your sins, first you added 100s of licensees and now you make consolidation more difficult,” says Uppal.

Remember, scale is what makes or breaks consolidation attempts, typically achieved when larger players merge or acquire each other. “I don’t know what Loop, Datacom bring to the table. If they don’t get acquired, why do they stick around? It is extremely late in the game,” says Pai of Tonse Telecom.

The question is, why would anyone want to acquire them? They have too few subscribers and they can’t be acquired for spectrum as most of it would have to be returned. “What are the benefits from M&A for Airtel? What would it get by, say, buying a new entrant?” asks Shetty.

Now the regulators may think that the small players will band together and form a large entity and go on to compete with the big boys but that’s pipedream, not policy. Those sort of mergers will not happen because one, they are too tough from a managerial point of view and two, even the new entrants want very high valuations. Many of them expect more than $100 per subscriber, regardless of whether she is real or a paper projection.

The reality meanwhile is that many of today’s subscribers will not generate that much during their entire lifecycle. “Valuations are still based from the Hutch acquisition days,” says Shetty. The post 3G valuation has Gupta of Nomura puzzled. “Will they [newer operators] demand 3G prices? Why would a company pay $3-4 billion for 3-4 million subs [subscribers]?” Some are even more drastic. “Subscriber-based valuation should simply be thrown out of the window,” says Barasia of Bain.

Is a Way Out Possible?
There are two or three ways out of this mess. One is to clearly have a market mechanism to price spectrum. Have frequent auctions to sell off whenever it becomes available. The other solution will be to create a spectrum trading market. It works for the electricity sector. The government decides an upper limit to keep speculation down, but no lower limit.

If this was introduced for spectrum it would allow companies that have excess of it to trade it and profit from it. The government can tax that profit if it wants to. If that’s too radical, allow all operators to share spectrum, leaving them and their equipment providers to work out the technical aspects around that?

The second way to make a sector stronger is to do away with the 30 percent market share limit in case of a merger or acquisition. There are many who advocate a pan-India limit of 40 percent and a 30 percent limit in each circle. But why have those caps in the first place. “The government has created the Competition Commission of India. Let it decide if the merged entity is detrimental to consumer interests,” says a senior telecom executive at one of the top three firms.

And lastly, don’t force companies to give away spectrum when they merge. Do two petrochemical companies have to return land when they merge? Do cement companies have to do it? Steel companies don’t shut down one plant facility and return it to the government when they merge. The whole idea is absurd.

And even more absurd is the idea of taking away the 900 MHz spectrum that was allocated 10 years ago and which is anyway a part of the global frequency plan of mobile systems. In any case, taking away the very bedrock on which all Indian mobile networks are based is in all probability legally not enforceable.

The government should try something like the US FCC did when it took an old frequency band in 700 MHz that television had vacated and auctioned it for mobile purposes.

Some of these modifications may come true but will take time. In the meanwhile, telecom companies will have to figure out new business models to compete and last the next 24 months or so.

New Ways to Compete
Consider even the market leader Bharti, India’s largest mobile operator with over 130 million subscribers. The Rs. 12,300 crore it spent on buying 3G spectrum when divided to a more meaningful metric means it paid Rs.1,471 for each of the 84 million subscribers across 13 circles.

Add to that the money it must now spend on upgrading its network to support 3G services — around $750 million say experts — and the cost per subscriber shoots up to Rs. 1,872.

 

Image: Sunil Mittal by Dinesh Krishnan, Anil Ambani by Sameer Joshi / Fotocorp; Sanjeev Aga by Sameer Joshi / Fotocorp

Which means either each of its subscribers in the 3G circles needs to bring in an additional Rs. 1,872, or a smaller number of higher-value ones even more, over the years for Airtel to just break even on 3G costs alone. Gupta of Nomura thinks it unlikely that in a price-sensitive mass market like India everyone will go in for 3G to begin with. The likely number, he feels, might be 15 percent. If that is the case, then Airtel will need to make an additional Rs. 12,486 from its top 15 percent customers to break even. That’s a lot of additional revenues.

The optimists say this money can easily come in with a negligible rise in ARPU over the 20 year license period for 3G spectrum. But that’s just a theoretical concept because no one, certainly not Bharti or Vodafone, will want to wait 20 years or even 10 years to recoup their 3G investments.

How long will it take then?
Though most experts agree that five to seven years is what it will take most operators to start breaking even on their 3G investments, the next two-three years will be the military bootcamp where the men get separated from the boys.

While the boys try peddling “minutes” to marginal customers at lower and lower prices, the men will be learning the ropes of the data services game. They will realise, for instance, that subscribers in small town India are just as interested in 3G as in Mumbai or Delhi, provided there is content they desire.

Fifty percent of the users viewing cricket IPL-related videos earlier this year came from “tier-2, tier-3 and tier-4 towns”, says Rajesh Reddy, CEO of mobile software maker, July Systems, which powered the mobile version for the tournament. Most of them used entry-level mobile phones made by the likes of Micromax or Lava, says Reddy based on his analysis of their browser details. And on an average each one watched two videos every day, “That’s higher usage than what we see for viewers of the NBA championships in the US!” he adds.

 

 Image: Arvind Rao by Gireesh G V, Neeraj Roy by Dinesh Krishnan; JS Sharma by Amit Verma; Nandan Nilekani by Mallikarjun Katakol

This means figuring out services that people would want to pay for. With voice, that was a no-brainer. Now companies will have to figure out location-based services. For instance, it could be that when you are in the vicinity of a mall you could be offered an instant time-bound discount through your mobile. “The mobile companies could get a share of revenues from malls or retailers for being a conduit for driving traffic into the outlet,” says a telecom industry executive.

Such content cannot come from operators, not just because they don’t have the money or expertise to build it, but because of the rise of a new class of mobile phones: The smartphones. Smartphones allow users to access a slew of services by either installing third party applications (apps) or through an Internet browser.

The place where mobile users head to check out their favourite apps is the “appstore”, a place where any developer can peddle his wares to any buyer with a phone. Seeing this trend, companies like Airtel, Aircel and OnMobile have launched their own appstores in India, apart from ones from mobile handset makers like Apple, Google Android or Nokia. The customer will have a lot to choose from.

Even the content creators like Hungama, Mauj and Indiagames will get more choice. To reach subscribers, most of them were at the mercy of operators who kept 70 paisa of every rupee they earned. This is the exact opposite of US and Europe where content creators get to keep 70 percent of what the customer pays. “When it comes to normalising their revenue share with content creators, operators will have no choice but to, because consumers will have choice now,” says Reddy.

KPMG’s Shetty says once operators clear the trial-by-fire of the next two-three years, India will be one of the most attractive mobile markets in the world. One of the biggest attractions will be mobile banking which he feels RBI might allow by that time, after testing the waters over the next few years. In addition the UID project will equip between 600-700 million Indians with something that was either missing or hard to prove: Their identity. That will create a new ecosystem of secure mobile transactions in areas like governance, banking and even telemedicine.

Meanwhile operators can, of course, sell the low hanging fruit of 3G dongles that can be plugged into computers to access the Net.

But many experts are sceptical of actual possible speeds through these dongles once a significant chunk of the 5 MHz spectrum — already much lower than the 10-15 MHz that is optimal — is carved away for carrying voice. Nonetheless, more of those will be launched by auction winners. But pricing on that will always be on a downward curve as has been the trend with Net tariffs.

But these things are untested and untried. The Indian consumer is yet to show an inclination to pay for data and data-based services. But a telecom firm really has no choice.

“In the future, telecom firms will have to do much more product development and segment the market according to price and behaviour than they have done before,” says Parida of Vodafone. Tonse Telecom’s Pai is much more blunt: “The days of jumping into the Indian market, adding subscribers and becoming EBITDA positive by the third year are gone for good.”

The Camelot of selling voice calls at high profits is over. The rigours of civilisation lie in wait.