Welcome to ZenUnwired; a blog dedicated to tracking developments in technology and strategy, and to deciphering the impact of these developments on wired and wireless ISP's, device manufacturers, OS and application developers, and most importantly - you.

Friday, March 26, 2010

The challenges of forecasting Indian growth

India is one of the most challenging mobile markets to analyse. Firstly because of its gargantuan size, secondly because of its sub-regional diversity, and last but not least because of the emergence of new regulations.

India is the second largest mobile market in the world and is bigger than any other region (apart from Asia Pacific of course); it recently reached 525 million connections which makes it twice the size of the USA and as big as Western Europe. Wireless Intelligence's latest set of connection forecasts show that India is playing catch up with China and there’s a very high probability that in just over five years it might become the world’s largest mobile market. That said, it is very challenging to predict how the Indian market will behave in the coming years.

The tremendous growth that the country has registered over the past few years and recent quarters is linked to the phenomenon of multiple connections per user. It is well-known that users tend to have more than one SIM card (my gut feeling would be an average of four per user) that they select depending on the territory they travel to or, more commonly, the latest cheapest prepaid deal on offer (85% of the user base in India is on prepaid tariffs). If tomorrow the regulator introduced a new regulation – like in Vietnam – limiting the number of SIM cards per user and asking all operators to remove inactive SIM cards from their subscriber base to report active subscribers, we would see operators’ figures shrink dramatically and market shares would be redistributed, to some extent, reflecting a very different picture.

This phenomenon of multiple connections per user is today distorting the market seasonality. Well, actually, I believe that the way Indian operators are reporting is hiding all sense of seasonality in the market. Take BSNL in Rajasthan as an example; in 1Q09 they reported quarterly net additions of 306,940, then reported a loss of 22,347 the following quarter but went back up by a staggering 324,940 net additions in 3Q09. Such a bumpy ride is often linked to either a) the launch of limited prepaid offers, b) the closure of part of their networks in the area, or c) the increase in network coverage. And we’ve got plenty of examples of operators which recently reported odd fluctuations in quarterly performances within several circles in India.

On top of that, not only is seasonality non-existent, but in some circles market penetration is meaningless. Since we’ve broken out our forecasts by circles, we’ve done a quick analysis on demographics using official government figures. The aim was to identify the potential population that mobile operators could target per circle – notably when launching 3G. So we came up with a basic equation that takes into account literacy and poverty levels, GDP per capita, number of students and workers, the share of rural populations, and so on. We have streamlined population figures to a more granular level which we can therefore use as a threshold to estimate the logical maximum penetration per circle. But the current market dynamics (as explained above) make it almost irrelevant at the moment. For instance, in the Metros, the market penetration has recently reached 137% in Mumbai, 294% in Kolkata, 151% in Delhi and some 263% in Chennai. But our streamlined population figures reflect peaks of 267% penetration in Mumbai, 556% in Kolkata, 295% in Delhi and 497% in Chennai! Obviously, both sets of penetration figures are our own estimates and are being used cautiously by our analysts to inform market growth scenarios, but even in Metros they can almost be dismissed when forecasting market growth as they do not reflect any logical maximum. Other factors are therefore taken into account to accurately forecast from market share assumptions to circles and regional benchmark.

Nonetheless, even though competition is intense in Metros, more operators are still launching their services there and investing money to cover the most penetrated part of India. I consider any new commercial launches in Metros as brave moves but if I were an investor I would rather focus on rural areas since that is where organic growth is coming from. Our latest set of forecasts clearly shows that the growth in Circles A, B and C are greater than the growth in Metros. Hence, we applauded the recent launch of new entrants, Uninor and Stel, in areas such as Bihar, Uttar Pradesh (West and East) or Orissa. Having said that though, some circles such as North East India are showing rather low penetration rate at present, but its growth rate is amongst the lowest in Circle C. There’s therefore plenty of room for growth in some areas but the pace of adoption highly depends on network roll-outs and regulations.

The role that AUSPI, COAI, TRAI and DoT play within the mobile industry in India is crucial. Local authorities have a huge task to monitor the growth of the industry and regulate it in line with the country’s economic and political priorities. Mobile number portability (MNP) is one of the new regulations which is likely to strongly impact mobile operator performances – in the same way it affected the Brazilian market. But MNP is following the 3G curse, as it is being delayed for security reasons, and has been pushed back again to May 2010. Another regulation already in place which was meant to filter market growth is the subscriber verification norm. But the DoT had to fine some operators in December last year for failing to adhere to it as they continued to register SIM cards under fake identities - especially in the northern circles of Jammu & Kashmir and Assam. One last example of the direct influence the regulator has on market growth can be found in recent headlines, as we read last month that local authorities in the city of Noida (Uttar Pradesh) shut down a quarter of the base stations in the city (some 190 towers) as operators installed them without planning permits. Finally, with the upcoming likely consolidation amongst operators, regulators have probably no time to rest.

All this shows the level of granularity we have to immerse ourselves into when we forecast mobile growth in India. The trends and market dynamics listed above are just the icing on the cake though; more specific factors are impacting the market and its potential over the coming years. Now that 3G is slowly but surely being squeezed into the picture, I believe that anyone failing to assess India using a bottom-up approach - at operator and circle levels - is very likely to be shortchanged.