Pot of gold or winner's curse? An event study of the auctions of 3G mobile telephone licences in the UK

Abstract

Commentators have suggested that the winning companies in the UK 3G mobile telephone auction overpaid for their licences. However, event-study method using the market model under ordinary least squares (OLS), robust and structural time-series estimation yields no systematic evidence of the ?winner's curse?. Positive as well as negative one-day wealth effects are observed amongst both winners and losers, and there is no lasting adverse market reaction to the winners, taken as a group. We conclude there is no case for easing the regulatory stance in the industry on grounds that the winners paid too much.Peer reviewe

Identifying the Winner's Curse in the First Spectrum Auction in the Republic of Korea Using an Event Study Approach

“…Despite previous studies not finding empirical evidence to support these negative effects of spectrum auctions in the mobile communications market [14], [15], doubts concerning the overbidding of mobile operators still remain after some European 3G auctions ended with unexpectedly high prices. In particular, a major concern over spectrum auctions is that the up-front license fees might lead to a winner's curse [16], a phenomenon whereby the winning bid is greater than the true value of the spectrum asset [4], [8]. Whether a spectrum auction can suffer from a winner's curse is impossible to say with theory alone; the identification of a winner's curse should be an empirical matter [4].…”

“…However, only a few previous studies have attempted to empirically investigate whether the winning companies experienced a winner's curse in some specific 3G auctions, such as those in the UK and Germany, in which unexpectedly high license fees were observed. First, the possibility of a winner's curse in the UK's 3G spectrum auction in 2000 was examined using an event study method [8] and option pricing technique [9], but there was no systematic evidence of a winner's curse in the auction. Second, using the same event study approach, in Germany's 3G spectrum auction in 2000, Mackley [4] empirically found that there was at least short-term evidence of a winner's curse.…”

“…An interesting point in the literature is that there is no consistency about whether a winner's curse existed in the spectrum auctions. There have been strong suggestions that the winning companies overpaid in both auctions [8], but they found different results regarding the existence of a winner's curse. This implies that the existence of a winner's curse in a specific region or country can be affected by the specific regional or national context.…”

“…called "winner's curse," a phenomenon whereby the winning bid is greater than the true value of the asset [4], [8] in spectrum auctions. Previous studies have investigated whether winning companies experienced such a curse in the European 3G spectrum auctions [4], [8], [9]. If a winner's curse occurred during the Republic of Korea's first spectrum auction, it would indicate that the government needs to rectify its forthcoming auction design.…”

“…This study may be the first attempt to examine the auction results for a winner's curse in a Korean auction. Although previous literature has examined the existence of a winner's curse in European 3G auctions using an event study approach [4], [8] and real option technique [9], the results might not be sufficient for the identification of a winner's curse in the first Korean auction, since they were limited to 3G auctions in countries such as the United Kingdom (UK) and Germany. Because the Korean case corresponds to the spectrum auction for advanced mobile communications services (4G) including Long Term Evolution (LTE), our results may provide a better understanding of the value of the 1.8-GHz spectrum in the Republic of Korea with policy implications.…”