Spectrum Auctions: The India Context

On 26 July 2022, the Department of Telecommunications (DoT) conducted the Spectrum auctions to allocate spectrum bands to telecom operators in the country. The spectrum comprises of airwaves/frequencies that are used in radio and cellular transmission and is divided into multiple bands based on these frequencies. The auction was an important event for the limited players it catered to, especially in offering 5G services in the circles they bid for. This year’s auction also saw a new entrant apart from the existing trio of Jio, Vodafone-Idea and Airtel – Gautam Adani’s Adani Data Networks. The structure and outcome of this auction will thus give us an insight into the industry’s future prospects. This article tries to determine whether or not such an arrangement is efficient and, if it is, what the payoffs for the players involved might be.

Why Auctions?

Before we dive into the world of spectrum auctions, it is imperative that we understand the science behind designing an auction. This will help in answering the question of efficiency; why the current format is or isn’t the right one for spectrum allocations.

Since we are dealing with a category as specific as spectrum auctions, the right auction format should be applied to achieve the objectives. For example, the English (increasing bids) and Dutch (decreasing bids) formats are synonymous with ‘private values’ for ‘single objects’. These formats can be labelled efficient since the objective of the highest bidder (with the highest willingness to pay) being rewarded is achieved. Price discovery for single-object auctions is thus feasible. But what if all the players ultimately share a common value for the object/s of interest? Irrespective of bid amounts, let’s say that the benefit derived from the ownership of the object is fixed for all. With multiple bidders and different information sets for each, it is likely that players will bid higher than the object’s common value. The winner, unfortunately, ends up overpaying and falls prey to what economists call the Winner’s Curse.

The only way to avoid this curse is to have perfect information on the object’s common value, which in the case of natural resources (in terms of potential production value) is highly uncertain. This is true with spectrum allocations too, given the uncertainty surrounding potential revenues from ownership of spectra.

Another problem with spectrum allocations is the presence of multiple bands in multiple locations. In the case of conventional auctions, bidders are required to value a single common object. However, a spectrum has multiple bands with varied frequencies, that too in different geographical locations. Sometimes, it is beneficial for telecom operators to hold multiple such bands. Say you desire a radio band in the Karnataka circle. But owning it would only make economic sense if you also owned television broadcasting bands in Kerala. Operators find advantage in providing a bundle of services rather than in discrete offerings. In a sequential auction, you never know whether both these objects will be within reach, and the uncertainty takes a toll on your bids.

The SMRA Model

Both the aforementioned problems combine to prevent players from making confident bids on individual items. The initial solutions were beauty contests and/or lotteries to allocated frequency bands. Such mechanisms allowed for spectrum allocations to a single or few operators based on a combination of operational capabilities and financial prowess, but not competitive bidding. However, this was far from optimal and also unfair to potential competitors. A path-breaking solution to this dilemma was provided in the early 90s by Nobel Prize-winning economists Robert Wilson and Paul Milgrom through their study of auction theory and design. Their proposal was a Simultaneous Multiple Round Auction (SMRA) format. In such an auction, any bidder can bid on any number of items (here, bands) at a time since multiple items are auctioned simultaneously. The second differentiating aspect is the multi-round mechanism, which allows bidders to revise their bids or place new bids for items in subsequent rounds. In the case of spectrum auctions, players can place bids on multiple bands and are required to place bids every round to avoid disqualification. Further, if for a particular band, no new bid is made in a round, the player with the current highest bid is awarded ownership of that band. While SMRA ensures a no-monopoly outcome, it also leads to efficient auctions where bidders can attain the right combination of bands given their willingness to pay. This format was first incorporated by the Federal Communications Commission of the USA in 1994 to allocate broadcasting licenses. It soon became the dominant model for radio frequency sales worldwide. Even Indian telecom authorities introduced this format in the late 2000s after an unsuccessful attempt with sealed-price auctions

India conducted its first SMRA auction for airwaves in 2010. As the name suggests, two distinct sub-auctions for the 2300 MHz (3G) and 2100 MHz (BWA) bands were held simultaneously. An open auction under the SMRA model, reduces the uncertainty over the valuation and thereby, reduces the chance of occurrence of winner’s curse. In the SMRA auctions, the reserve price is the minimum opening bid price. So, setting a reasonable reserve price is crucial to the success of the auctions and, eventually, the quality of the services provided.

Economic Implications of High Reserve Price

About 71% of the spectrum on offer secured bids in the recently held spectrum auctions. The Government garnered a little more than ₹1.5 lakh crore in revenue, which is just 35% of the reserve price it had set for the spectrum on offer. After going unsold in the last two auctions, only 40% of the 700 MHz band secured bids. This band is crucial for extended coverage and deeper penetration. Almost all the other bands got sold at reserve prices, and the 600 MHz bands went unsold. According to an analysis of the SMRA auctions for the 700 MHz band held worldwide by the EPW, the average reserve price per MHz per population is $0.05 (adjusted for PPP). The mean winning bid price was $0.54, which indicates a good price discovery. Therefore, a major cause for concern is the reserve pricing. Exorbitant reserve prices would force the bidders to miss out on crucial bands, which are essential to creating an efficient network. High reserve prices might lead to frivolous bidding, encourage collusion and speculation and make the recovery of the administrative costs difficult. The sale price approximately being equal to the reserve price also hinders the price discovery during the auction process. It might lead to the underutilization of the spectrum available for sale resulting in higher costs for the consumers and poor quality of services.

Despite its benefits, the SMRA model also exposes the bidders to aggregation risk. Aggregation risk happens when the bidders miss out on bidding for a key band, essential to create an efficient network. India’s telecom market is divided into 22 circles and telecom operators often do not get contiguous bands of spectrum due to various reasons. Being overcautious of the winners curse, although it is non-existent in the Indian context, is one of the reasons for such underutilisation.

Thus, aggregation risk and high reserve prices are the main threats to SMRA spectrum auctions in India.

The Way Ahead

Given that the ARPU (Average Revenue Per User) of the Indian Telecom industry is among the lowest in the world, high reserve prices would cripple the operators with high debt. The best way forward is to bring the reserve price in line with the global average in subsequent auctions and not just by taking the last winning price as a reference. In order to reap all the benefits of the 5G services, spectrum allocation should be done across all ranges of frequencies – low, mid and high. Finally, each operator should be allocated a certain degree of contiguous spectrum to ensure efficient mobile internet connectivity.









Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s