How the spectrum game is played

Frequent call drops and poor mobile broadband connections are a way of life. Such is the public outcry, frustration and fury about it that the government has drafted guidelines to punish mobile operators for poor network connectivity. Service providers (or operators), on the other hand, are up in arms holding the policy of limited spectrum allocation responsible for the poor service.

An obvious solution to the spectrum scarcity problem is for the government to release and provide more unused spectrum. Alternatively, improvement in spectrum technology could allow for more efficient use of available spectrum. Besides these, there are at least two other ways of addressing the problem of poor service without increasing spectrum supply or waiting for technological innovations. First, operators could agree to merge with or acquire other providers. Second, operators could trade or share existing spectrum using secondary spectrum markets.

Mergers and acquisitions are a useful market-based tool for spectrum reallocation. Though the total quantity of spectrum during an M&A will not change, possible under-utilised spectrum held by one of the firms will be put to better use and trunking gains will improve spectral efficiencies. Of course, since M&A involves transfer of networks and subscribers apart from spectrum, it is likely to happen only when firms want clear exits.

Secondary spectrum markets, much like similar markets in other contexts, including automobiles, stocks and housing, facilitate resource re-allocation. The recently released trading and sharing guidelines for radio spectrum provide an opportunity for operators with very high MHz per million subscribers, to offload under-utilised spectrum through the secondary market.

In either of the two routes to spectrum re-allocation, regulatory guidelines impose significant restrictions, ostensibly to ameliorate concerns about monopoly power and abuse of dominance. The first of these relate to spectrum caps — limits on the amount of spectrum an operator may hold.

Within each circle, no operator can hold more than 25 per cent of total spectrum holdings nor more than 50 per cent of the spectrum in each band. Additionally, M&A guidelines issued last year impose another cap of 50 per cent of market share (both in subscriber base and adjusted gross revenue) for the merged entity in each circle.

It is possible that if operators are at these limits, there would be little room for either secondary market or M&A activity. In an M&A situation, the guidelines require the entity to ‘surrender’ the excess spectrum within one year of the transaction, with no refund! However, it is easy to get around the no-refund policy. If there is an active secondary market, it would be rational and feasible for the merging firms to sell the excess spectrum prior to consummating the M&A transaction.

We take a quick look at the existing spectrum holdings across different circles to evaluate how close the different operators are to the implied spectrum caps. Besides a few operators in some bands in a handful of circles such as Maharashtra, Punjab, Bihar, the North-East and Odisha, the spectrum cap is not binding. The maximum market share of an operator in any of the circles is about 33 per cent, leaving significant headroom for M&A. In fact, most of the newer operators are struggling to get subscribers and have extremely high MHz per million subscriber values. These operators could be potential participants in M&A deals.

While the spectrum caps are not immediately binding, such constraints need to be rationalised to ensure they do not hinder the market mechanics of efficient spectrum allocation. If it were possible for operators to swap (a two-way sale) as against a one-way sale, many more transactions may become feasible.

Alternatively, operators could choose to ‘share’ spectrum since they are deemed to hold only 50 per cent of shared spectrum for spectrum cap calculations. Besides, sharing does not involve the risk of losing the golden goose (radio spectrum) while at the same time managing supply and demand for optimal utilisation of spectrum by contracting parties. Spectrum sharing also has the advantage that it involves contracts of lower time periods, giving enough flexibility to the parties involved. Of course, sharing involves paying an additional 0.5 percent of revenue as annual spectrum usage charge.

Sharing spectrum is not without regulatory constraints. Unlike spectrum trading, sharing is restricted to operators within the same circle and who have spectrum within the same band. Under this restriction, for instance, Aircel will be unable to share in the 800MHz band with another operator since it currently has no spectrum in that band in any circle. For the same reason, Aircel would also have no sharing opportunity in the 900 MHz band in Mumbai. These restrictions limit the feasible set of sharing partners, thereby hindering the functioning of the secondary market.

Along similar lines, the guidelines do not permit sale of spectrum for part of a licensed service area. This is particularly restrictive in a country where there are large variations across districts. It is possible that potential buyers have district-specific advantages in service delivery. If the burden of providing service in all districts is particularly onerous, an interested buyer with an advantage in one district but not in others would likely remain out of the market completely.

Finally, the elephant in the room is the treatment of non-liberalised spectrum for secondary market transactions. The spectrum acquired administratively and not yet liberalised through payment of one-time fee dating back to 2008 can be neither traded nor shared. This takes out about 50 per cent of the assigned spectrum blocks in 800 MHz and 1800 MHz from the secondary market. The current guidelines of paying one-time fee to liberalise spectrum may not be economically attractive for license holders, thus restricting spectrum trading or sharing.

In summary, the guidelines allowing trading or sharing of spectrum impose constraints that may undo or, at the very least, undermine the attempt at efficient spectrum reallocation. Active secondary spectrum markets are likely to reduce industry costs, promote innovation and trigger diffusion of mobile services. Limited activity would be the death knell for a fledging secondary market in spectrum.

The writers are faculty members at the University of Maryland at College Park, MDI Gurgaon and IIIT Bengaluru respectively

COMMENT NOW