ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Fixing Appropriate Reserve Prices for 5G Radio Spectrum

This article analyses the spectrum-winning bid prices in simultaneous multiple round ascending auction held across the world for 700 MHz and indicates how it had an impact on the winning bid prices in other countries. It also analyses the spectrum prices of the mid-band (3.3–3.6 GHz) and the millimetre wave band (26 GHz) that are essential for the rollout of 5G services and makes a similar remark on reserve prices. Based on the analysis, it predicts the possible outcome in the forthcoming spectrum auction.

The fifth-generation (5G) communication services are being rolled out worldwide with 135 commercial launches covering about 235 million subscribers. It is estimated that 5G services will cover 3 billion users by 2025. Apart from the popular views about 5G providing high-speed data transfers, it is more about bringing a paradigm shift in ways in which things and humans are interconnected. The McKinsey Global Institute (MGI 2020) has estimated an increase in global gross domestic product (GDP) by about $2 trillion through the use of the 5G connectivity paradigm in key sectors such as healthcare, retail, mobility, and manufacturing alone.

In India, we have made remarkable inroads by covering close to 800 million subscribers with 4G network services. In order to cover the rest of the population, including rural areas with 4G/5G, it is imperative that the government make every effort to sell the radio spectrum, which is a critical resource in the forthcoming auctions. Since radio spectrum cannot be stored for future use, it is important to make effective and efficient use of it. By not being able to utilise about 1,200 megahertz (MHz) of the 700 MHz spectrum since 2016, we have lost a huge value of this scarce resource.

In this article, we analyse the worldwide trends in the prices of two spectrum bands that are most likely to be put into auction soon in India. First, we look at 700 MHz, the low-frequency spectrum that provides coverage across wide areas—the “coverage spectrum.” Subsequently, we analyse 3.3–3.6 gigahertz (GHz) and 2,628 GHz—the “capacity spectrum” that provides a high-capacity in dense urban areas.

The Sub-GHz Band

The 700 MHz radio spectrum went unsold for the second time during the last spectrum auction in February 2021; the first such failure was in 2016. This sub-MHz radio spectrum is normally considered as the “golden band” as it has superior propagation characteristics, with good penetration inside buildings and is also suitable for both rural and urban coverage. Many countries, including Europe, have prioritised the allocation of 700 MHz for 5G services. The GSMA (2019), the industry body of mobile operators around the world, has specified that the number of sites used in 2,500 MHz (that has been auctioned in India recently) is seven to eight times more than that required in 700 MHz for covering the same area, thereby indicating the superior coverage capabilities of this spectrum band. Given all these benefits, it begs a question as to why no bidder picked up this band during the last two auctions. Is the culprit the reserve price set for this band in these auctions?

Reserve price is the opening minimum bid price in simultaneous multiple-round ascending (SMRA) auctions and setting it appropriately determines whether the auction will be successful or not. Setting too high a reserve price will dissuade the bidders from bidding as it might exceed their valuation of the spectrum; too low a reserve price might be seen as the auctioneer giving away the scarce spectrum at a minimal price.

We analysed the SMRA auctions held worldwide for the 700 MHz spectrum using the spectrum database of DotEcon (2020). The mean reserve price per MHz per population is about $0.05 (in dollars adjusted for purchasing power parity [PPP]). However, the mean winning bid price is about $0.54. The average winning bid price is about $0.33 per MHz per population in the United States (US) that conducted a number of SMRA auctions for 700 MHz during 2002–11. Figure 1 (p 21) provides the variations in winning prices of 700 MHz bands auctioned across different countries in various auctions conducted over the years.

On the contrary, the average reserve price fixed by the Government of India for 700 MHz in the forthcoming auction is about $0.24 per MHz per population (adjusted for PPP), that is several times higher than that set internationally and is also higher than the mean winning bid price witnessed worldwide. A closer analysis indicates that the reserve price in metros is about $1, around twice the mean winning bid prices in other countries. Only the category C circles had the reserve price set at $0.08, which is a bit closer to the mean reserve price in other countries. Though the reserve price set in the 2022 auction is much lower compared to that fixed in the 2016 and 2021 auctions, it is still very high for the operators to place their bids in the first clock stage. The same is true for the newly released 600 MHz spectrum.

There have been many studies on the determinants of winning bid prices in auctions, the notable one being Milgrom et al (2017). The ordinary least square (OLS) regression of the winning price with the explanatory variables such as (i) the reserve price; (ii) the population covered in the spectrum licence area; (iii) the total amount of available spectrum put on auction; (iv) the total amount of allocated spectrum in the auction; (v) the number of bidders in the auction; and (vi) the duration of spectrum licence, as in Sridhar and Prasad (2021), indicate a strong positive correlation between the reserve price and the winning price. Results of the regression are presented in Table 1.

As indicated, for every 1% increase in the reserve price, there is a corresponding increase of almost 1% in the winning spectrum prices. However, it must be pointed out that the lower reserve prices do not necessarily mean lower winning prices. For example, in the 2011 auction conducted in the US, the average winning bid price was 10 times more than the reserve price. Even Germany witnessed more than two times the reserve price as the winning price in the auction held in 2015. However, Finland that conducted the auction for 700 MHz in 2016 witnessed only marginal increase in the winning price. Hence, it all depends on how the bidders value the spectrum (Sridhar and Prasad 2021).

The Mid Band

We also analysed worldwide prices of the spectrum in the 3.3–3.6 GHz range that is slated for auction soon. This spectrum band is ideal for providing spot coverage of high capacities and ideally suited for dense urban areas of the country. Thus far, nine countries have auctioned this spectrum using SMRA for mobile services and more than 30 operators are providing 5G services in this band. Figure 2 provides an illustration of the spread of the winning bid prices across countries over time. There is a trend towards an increase in prices as 5G networks get rolled out by the carriers in different countries and the consequent effects on demand for spectrum.

The mean reserve price per MHz per population or this band in auctions conducted worldwide is about $0.02 per MHz per population, while the mean winning bid price is about $0.11. As expected, the price of this high-frequency spectrum is about one-fourth of the low-frequency 700 MHz spectrum prices. The reason for this variation in spectrum prices between coverage and capacity spectrum is due to the trade-off the bidders make between spectrum and the mobile access infrastructure as illustrated by Prasad and Sridhar (2014).

The regression analysis of the awards of this mid-band spectrum, much similar to what is done for 700 MHz, do not indicate significant correlation between reserve price and winning price, with a few exceptions. In Germany, though the auctioneer had set a very low reserve price of $0.002, the winning price reached 100 times to about $0.24.

The Telecom Regulatory Authority of India (TRAI 2018) has recommended a mean reserve price for the mid band at $0.02, which is in-line with the mean reserve price worldwide. At metros, the reserve price, much similar to the 700 MHz, is high at $0.08. However, this is still less than the mean winning price in this band.

The Millimetre Wave Band

The millimetre wave band in the high frequencies of 26–28 GHz is ideal for creating micro and pico cellular deployment of 5G. Worldwide, about 52 countries have completed auctioning this band for 5G. The average reserve price recommended for this band is about $0.0004 per MHz per population, which is about the world average (Sridhar 2022). Overall, we see muted response to all bands in the forthcoming auction, except for 3.3 GHz and 26 GHz bands.

Policy Implications

Since 1994, when Nobel Prize laureates Paul R Milgrom and Robert B Wilson along with McAfee proposed to the Federal Communications Commission the conduct of the first-ever SMRA auction as the mechanism design for allocating radio spectrum for commercial mobile services in the US, it has been used successfully worldwide. However, as indicated, an unrealistic reserve price can upset the apple cart. In a recent cross-country study of spectrum prices, Sridhar and Prasad (2021) indicate that the reserve prices have a significant and positive effect on the winning bid prices.

The other factor that is important as a determinant of spectrum prices is the availability of spectrum. As indicated in Table 1, an increase of 0.8 kilohertz (kHz) of available spectrum decreases the spectrum price by 1%. Though the effect is less, at least, in the auction of 700 MHz, Lewis (2008)—in one of the recommendations—indicates that making more spectrum available is necessary to make spectrum auction successful. Prasad and Sridhar (2014) also argue extensively that a shortage of available spectrum is one of the important reasons for high prices witnessed in India. Since the quantum of spectrum in this auction is much larger (about 35 times more than in 2021), including the new bands in 600 MHz, 3.3 MHz and 26 GHz, we expect that it will have substantial negative effects on spectrum prices (Sridhar and Prasad 2022).

As Kirkpatrick (2021) points out, the mid-band spectrum is being used by various government agencies, including the Department of Defense (DoD) in the US and the Department of Space (DoS) in India. There have been hurdles in freeing up the unused/underutilised spectrum for commercial mobile services from the current spectrum holders. In India, talks are still going on between the Department of Telecommunications (DoT), DoS and the DoD for the release of unused and underutilised spectrum in the 3.3–3.6 GHz band. The parliamentary standing committee on information technology recommended expediting the deliberations for the earliest identification and release of adequate spectrum for 5G services (Lok Sabha Secretariat 2021).

India leapfrogged into digital 2G technologies and subsequently had exponential growth on mobile services, surpassing the US to be the second largest, next only to China in subscriber base in the world (Sridhar 2012). It is time that we migrate to 5G and not lag behind other countries and to reap the benefits of super-connectivity between humans, and devices.

References

DotEcon (2020): “Spectrum Awards Database,” www.dotecon.com.

GSMA (2019): “5G Spectrum: GSMA Public Policy Position,” GSM Association.

Kirkpatrick, K (2021): “Can You Hear Me Now That I’m On 5G? Communications of the ACM,”
viewed on 22 June 2021, https://cacm.acm.org/news/249963-can-you-hear-me-now-that-im-on-5g/fulltext#.

Lok Sabha Secretariat (2021): “The Twenty-first Report of the Standing Committee on Information Technology (2020–21) on the Subject of ‘India’s Preparedness for 5G’ Relating to the Ministry of Communications,” viewed on 15 May 2021, http://164.100.47.193/lsscommittee/Communications%20and%20Information%20Technology/pr_files/Press%20Release%20on%2021st%20Report%20.pdf.

Lewis, C (2018): “Lessons from Spectrum Auctions: A Benchmark Approach,” available at SSRN 3185752.

MGI (2020): “Connected World: An Evolution in Connectivity beyond the 5G Revolution,” McKinsey Global Institute.

Milgrom, Paul and Ilya Segal (2017): “Designing the US Incentive Auction,” Handbook of Spectrum Auction Design, Martin Bichler and Jacob Goeree (eds), Cambridge: Cambridge University Press, pp 803–12.

Prasad, R and V Sridhar (2014): The Dynamics of Spectrum Management: Legacy, Technology, and Economics, Oxford University Press.

Sridhar, V (2012): Telecom Revolution in India: Technology, Regulation and Policy, New Delhi, India: Oxford University Press.

— (2022): “Are Spectrum Reserve Prices Too High?” Business Line, 26 April.

Sridhar, V and Rohit Prasad (2021): “Analysis of Spectrum Pricing for Commercial Mobile Services: A Cross Country Study,” Telecommunications Policy, Vol 45, No 9, https://doi.org/10.1016/j.telpol.2021.102221.

— (2022): “Analysing Spectrum Auction,” Hindu, 22 June.

TRAI (2018): “Recommendations on Auction of Spectrum in 700 MHz, 800 MHz, 900 MHz, 1800 MHz, 2100 MHz, 2300 MHz, 2500 MHz, 3300-3400 MHz, 3400-3600 MHz Bands,” Telecom Regulatory Authority of India.

 

Dear Reader,

To continue reading, become a subscriber.

Explore our attractive subscription offers.

Click here

Updated On : 24th Jul, 2022
Back to Top