Analysis of the first frequency auctions in the Russian Federation
Over the past year, two auctions were held in Russia to sell licenses for the use of radio frequency spectra. Both auctions were organized by Roskomnadzor on the trading platform of Sberbank-AST, and became the first market mechanism for the allocation of the frequency spectrum (before that, the decision was made on the basis of a competition). The first of the two auctions took place on September 29, 2015, collecting revenue approximately six times the initial value of the lots. At the same time, 10 licenses in the bands 1710-1785 MHz and 1805-1880 MHz were distributed between four companies - MTS, MegaFon, VimpelCom and T2 Mobile. The next auction for the range of frequencies 2570-2595 MHz and 2595-2620 MHz was held in February 2016 and collected revenue only 40% higher than the initial cost of lots, which is significantly less than in the September auction. What is the reason for such a difference in the auction proceeds?
Despite the fact that the auction journal with detailed actions of participants on the Roskomnadzor website could not be found, an analysis of open sources of information makes it possible to hypothesize that the chosen auction scheme and weak competition are ineffective.
To begin, consider the first auction. All licenses in it can be attributed to one of three groups: four licenses of south-west Russia (Karachay-Cherkessia, Stavropol Territory, North Ossetia and Dagestan), four licenses of central Russia (Perm Territory and Komi-Perm District, Samara Region, Orenburg Region ) and two licenses of eastern Russia (Buryatia, Amur region). Such a classification is convenient, since the value of lots for most telecom companies depends, among other factors, on geographic location, as well as the neighborhood of the areas corresponding to lots. For example, companies usually value lots in neighboring Dagestan and the Stavropol Territory higher than non-adjacent Dagestan and Perm Territory. The reason for this may be many factors, such as the fact that having licenses from neighboring areas, it is easier for companies to solve the problem of signal interference and use a higher data transfer rate. Therefore, obtaining licenses from neighboring regions creates additional value for telecom companies. That is, if the value of the license of each of the regions separately, say, 1 million rubles, then the value of two adjacent areas is higher than 2 million rubles (i.e., the licenses in this case are complementary, or complementary).
However, the chosen auction scheme does not allow participants to express this added value through a bid. Indeed, the auction is arranged in such a way that each lot is played independently of other lots, and it is impossible, say, to state a high bid immediately for a couple of areas corresponding to our example to Dagestan and the Stavropol Territory, and a low bid for each of these areas separately. This restriction creates additional risks for the telecom company: an auction participant whose business plan is to deploy a network in a particular region, such as the North Caucasus, must declare a high bid for each of the regions in the region. At the same time, having lost a single lot, for example, of the Stavropol Territory, the business plan may already become impracticable and, as a result, the value of already acquired lots of the neighboring areas of this region for this participant falls. Such risks usually lead to attempts by players to manipulate the auction and, consequently, to a low market efficiency.
')
Let's look now at the result of the September auction. We see that the MTS, having won in the Perm Territory (including the Komi-Permyatsky District) and the Orenburg Region, is most interested in the frequencies of central Russia. At the same time, it is logical to assume that MTS actively fought for the license of the Samara Region, although it did not receive it. This assumption is supported by the fact that the final rate for the Samara Region is almost 10 times higher than the initial one, which indicates an intense competition for this area. In addition, MTS already has a license in the neighboring Orenburg region and obtaining a complementary license from the Samara region would add value to the lot already won by MTS. We also note that the total cost of a lot in the Orienburg Region is about eight times higher than the initial one, while the license cost of the Perm Region is only four times higher than the initial value. This indirectly confirms the assumption that the neighboring Samara and Orienburg regions do have a higher value and caused the most fierce competition for a license than the isolated Perm region. Similarly, the initial price increase in other non-adjacent regions, such as Buryatia and the Amur region, was about four times.
The situation is similar in the south-western group: here, the largest increase in the initial price of the lot was experienced by the Stavropol Territory, reaching VimpelCom. It can be assumed that in this case the struggle for the license of this region was fought with MegaFon, who won the licenses in two neighboring regions (Dagestan and Karachay-Cherkessia).
Now let's look at the results of the second auction, held in February 2016. MTS won the federal lot, did not participate in the rest of the auction, which in itself reduces the competition, leaving only two major players. From the minutes of the meeting of the commission for the auction on the results of the consideration of applications it follows that for most of the lots the struggle was conducted only between these two players. Obviously, having the experience of the first auction, the players themselves, in order to avoid the risks described above, are more profitable to agree on who gets what area and declare the appropriate bids. This will result in everyone receiving part of the spectrum in adjacent areas, which will allow to realize the added value of the lots won by obtaining complementary licenses and, moreover, will eliminate the need to significantly increase the initial price of the lot. Now, looking at the results of the auction, we see that of the 40 regions that MegaFon received 35, indeed, form a single cluster in which each of the won regions is adjacent to the other region won by this player (with the exception of Ingushetia and North Ossetia, then Rostov, Belgorod and Oryol regions). Although there is no real evidence of collusion, it’s surprising that in this auction we don’t see the same struggle between MegaFon and VimpelCom as the one that was over the Stavropol Territory in the September auction (and which prevented MegaFon from getting a single cluster from Karachay-Cherkessia , Stavropol Territory and Dagestan).
Thus, in addition to a small number of participants, the problem of low revenues could be aggravated by an inefficient auction mechanism, which did not allow participants to completely “express” their business plan through appropriate bids for groups of areas instead of each separate area. This could lead to collusion or simply to the coordination of the actions of the participants with each other in the auction process and, as a result, to a significant drop in the auction proceeds.
The described problem can be relatively easily solved by Roskomnadzor (or rather, the Ministry of Communications and Mass Media) by changing the auction mechanism. For example, looking at the experience of countries such as the United States, the United Kingdom or Switzerland, which have long since allocated frequencies through an auction, one could use a combinatorial auction scheme, such as the one used by the Federal Communications Commission, FCC. In this combinatorial auction, market participants have the opportunity to more fully express their business plan through an arbitrary bid. For example, Megafon could declare that it estimates the licenses of three regions of Karachay-Cherkessia, the Stavropol Territory and Dagestan at once in 1500 million rubles and each of these areas in each 300 million rubles. This would significantly reduce risks and make it easier for telecom companies to participate in the auction. Licenses are then distributed optimally, maximizing social welfare, which, in the absence of market manipulation by players, leads to an effective outcome. Various modifications of such a combinatorial mechanism guarantee the absence of the creation of coalitions of participants, as well as minimize the possibility of market manipulation. This often makes them the standard mechanism for conducting frequency auctions in other European and North American countries.
The author is a doctoral student at the University of Zurich, Dmitry Moore
PS An appeal was received to the appeal to the Ministry of Communications and Mass Communications with a proposal to improve the auction mechanism that certain details, such as the prospects of each range, as well as the level of development and the amount of equipment already available, were not taken into account during the analysis, but On the whole, the approach described by the combinatorial mechanism is constructive in nature and will be considered further in the conduct of such auctions. According to the author, many details were indeed omitted in this analysis, but this was done, rather, because of the desire to illustrate the problem in a “first approximation”. Additional details make the nature of such auctions even more combinatorial and, as a result, motivate the change of the frequency auctions mechanism even more.
Despite the fact that the auction journal with detailed actions of participants on the Roskomnadzor website could not be found, an analysis of open sources of information makes it possible to hypothesize that the chosen auction scheme and weak competition are ineffective.
To begin, consider the first auction. All licenses in it can be attributed to one of three groups: four licenses of south-west Russia (Karachay-Cherkessia, Stavropol Territory, North Ossetia and Dagestan), four licenses of central Russia (Perm Territory and Komi-Perm District, Samara Region, Orenburg Region ) and two licenses of eastern Russia (Buryatia, Amur region). Such a classification is convenient, since the value of lots for most telecom companies depends, among other factors, on geographic location, as well as the neighborhood of the areas corresponding to lots. For example, companies usually value lots in neighboring Dagestan and the Stavropol Territory higher than non-adjacent Dagestan and Perm Territory. The reason for this may be many factors, such as the fact that having licenses from neighboring areas, it is easier for companies to solve the problem of signal interference and use a higher data transfer rate. Therefore, obtaining licenses from neighboring regions creates additional value for telecom companies. That is, if the value of the license of each of the regions separately, say, 1 million rubles, then the value of two adjacent areas is higher than 2 million rubles (i.e., the licenses in this case are complementary, or complementary).
However, the chosen auction scheme does not allow participants to express this added value through a bid. Indeed, the auction is arranged in such a way that each lot is played independently of other lots, and it is impossible, say, to state a high bid immediately for a couple of areas corresponding to our example to Dagestan and the Stavropol Territory, and a low bid for each of these areas separately. This restriction creates additional risks for the telecom company: an auction participant whose business plan is to deploy a network in a particular region, such as the North Caucasus, must declare a high bid for each of the regions in the region. At the same time, having lost a single lot, for example, of the Stavropol Territory, the business plan may already become impracticable and, as a result, the value of already acquired lots of the neighboring areas of this region for this participant falls. Such risks usually lead to attempts by players to manipulate the auction and, consequently, to a low market efficiency.
')
Let's look now at the result of the September auction. We see that the MTS, having won in the Perm Territory (including the Komi-Permyatsky District) and the Orenburg Region, is most interested in the frequencies of central Russia. At the same time, it is logical to assume that MTS actively fought for the license of the Samara Region, although it did not receive it. This assumption is supported by the fact that the final rate for the Samara Region is almost 10 times higher than the initial one, which indicates an intense competition for this area. In addition, MTS already has a license in the neighboring Orenburg region and obtaining a complementary license from the Samara region would add value to the lot already won by MTS. We also note that the total cost of a lot in the Orienburg Region is about eight times higher than the initial one, while the license cost of the Perm Region is only four times higher than the initial value. This indirectly confirms the assumption that the neighboring Samara and Orienburg regions do have a higher value and caused the most fierce competition for a license than the isolated Perm region. Similarly, the initial price increase in other non-adjacent regions, such as Buryatia and the Amur region, was about four times.
The situation is similar in the south-western group: here, the largest increase in the initial price of the lot was experienced by the Stavropol Territory, reaching VimpelCom. It can be assumed that in this case the struggle for the license of this region was fought with MegaFon, who won the licenses in two neighboring regions (Dagestan and Karachay-Cherkessia).
Now let's look at the results of the second auction, held in February 2016. MTS won the federal lot, did not participate in the rest of the auction, which in itself reduces the competition, leaving only two major players. From the minutes of the meeting of the commission for the auction on the results of the consideration of applications it follows that for most of the lots the struggle was conducted only between these two players. Obviously, having the experience of the first auction, the players themselves, in order to avoid the risks described above, are more profitable to agree on who gets what area and declare the appropriate bids. This will result in everyone receiving part of the spectrum in adjacent areas, which will allow to realize the added value of the lots won by obtaining complementary licenses and, moreover, will eliminate the need to significantly increase the initial price of the lot. Now, looking at the results of the auction, we see that of the 40 regions that MegaFon received 35, indeed, form a single cluster in which each of the won regions is adjacent to the other region won by this player (with the exception of Ingushetia and North Ossetia, then Rostov, Belgorod and Oryol regions). Although there is no real evidence of collusion, it’s surprising that in this auction we don’t see the same struggle between MegaFon and VimpelCom as the one that was over the Stavropol Territory in the September auction (and which prevented MegaFon from getting a single cluster from Karachay-Cherkessia , Stavropol Territory and Dagestan).
Thus, in addition to a small number of participants, the problem of low revenues could be aggravated by an inefficient auction mechanism, which did not allow participants to completely “express” their business plan through appropriate bids for groups of areas instead of each separate area. This could lead to collusion or simply to the coordination of the actions of the participants with each other in the auction process and, as a result, to a significant drop in the auction proceeds.
The described problem can be relatively easily solved by Roskomnadzor (or rather, the Ministry of Communications and Mass Media) by changing the auction mechanism. For example, looking at the experience of countries such as the United States, the United Kingdom or Switzerland, which have long since allocated frequencies through an auction, one could use a combinatorial auction scheme, such as the one used by the Federal Communications Commission, FCC. In this combinatorial auction, market participants have the opportunity to more fully express their business plan through an arbitrary bid. For example, Megafon could declare that it estimates the licenses of three regions of Karachay-Cherkessia, the Stavropol Territory and Dagestan at once in 1500 million rubles and each of these areas in each 300 million rubles. This would significantly reduce risks and make it easier for telecom companies to participate in the auction. Licenses are then distributed optimally, maximizing social welfare, which, in the absence of market manipulation by players, leads to an effective outcome. Various modifications of such a combinatorial mechanism guarantee the absence of the creation of coalitions of participants, as well as minimize the possibility of market manipulation. This often makes them the standard mechanism for conducting frequency auctions in other European and North American countries.
The author is a doctoral student at the University of Zurich, Dmitry Moore
PS An appeal was received to the appeal to the Ministry of Communications and Mass Communications with a proposal to improve the auction mechanism that certain details, such as the prospects of each range, as well as the level of development and the amount of equipment already available, were not taken into account during the analysis, but On the whole, the approach described by the combinatorial mechanism is constructive in nature and will be considered further in the conduct of such auctions. According to the author, many details were indeed omitted in this analysis, but this was done, rather, because of the desire to illustrate the problem in a “first approximation”. Additional details make the nature of such auctions even more combinatorial and, as a result, motivate the change of the frequency auctions mechanism even more.
Source: https://habr.com/ru/post/300520/
All Articles