Last week, the OFTA concluded its auction for spectrum in the 2.3GHz and 2.6GHz bands, with 2.6GHz paired spectrum being priced at HKD 2.44/MHz/pop, or EUR 0.24/MHz/pop. The credit crisis probably explained the lack of interest from foreign bidders, and the fact that competitors Hutchison and PCCW joined forces.Still, this outcome exceeds prices paid in other recent 2.6GHz auctions, in line with expectations. That said, unpaired spectrum seems to be a different matter altogether.
Comparing auction outcomes is always a tricky exercise. Many factors need to be considered: license requirements, auction design and local market conditions may vary significantly across auctions. Still, it can be argued that recent auctions, including the just-concluded auction in Hong Kong, seem to give quite a consistent picture of spectrum prices if such factors are taken into account.
The low outcome of the Norwegian auction in November 2007 (EUR 0.03/MHz/pop for paired spectrum) can largely be explained by the market structure in Norway. Incumbent operators are the ones who typically value spectrum the most. There are only two players in the Norwegian mobile market, so competition during the auction was relatively limited. This in turn causes prices to remain well below value. In this particular case, it even allowed Telenor to acquire part of the unpaired spectrum such that it could be used as paired, which effectively meant that there was more paired spectrum available than the original band plan indicated.
In the May 2008 auction in Sweden, a substantially higher price was paid for paired spectrum (EUR 0.16/MHz/pop). Given the fact that in Sweden four incumbents had to compete in the auction, and that collectively they owned only a relatively modest amount of spectrum before the auction, the higher price paid did not come as a surprise.
The Hong Kong auction has increased prices even further. Given the ongoing credit crunch and worsening economic outlook, this may seem surprising. However, there are good reasons for this:
- The most obvious one follows from Hong Kong's high population density, which leads to larger cost savings (on additional sites to cope with traffic) from additional spectrum on a per pop basis. Given that 2.6GHz is most likely to be deployed in urban areas only, it would not be fair to compare population densities on a nationwide basis. Still, cities like Stockholm and Oslo are less densely populated than Hong Kong, which justifies higher prices in the latter.
- This point was enforced by the fact that in Hong Kong, there are even more incumbents1 than in Sweden, competing for less paired spectrum. Moreover, this spectrum was split up in only three licenses, less than the number of incumbents competing in the auction.
Despite the achieved prices, it is likely that the credit crunch has affected the auction. The crunch has impacted the lending capability of operators, which in turn may explain the lack of interest from foreign bidders and the consortium formed by competitors Hutchison and PCCW. However, from an equity perspective, the situation is less dramatic. Although since the Swedish auction the Dow Jones World Index has dropped by 47%, telecoms stocks in the Asia-Pacific region have fallen by only 29%. Not an attractive picture either, but considerably less disastrous. Moreover, at the same time, the mobile data growth curve, which ultimately drives spectrum demand, seems to have gone in the opposite direction. For the moment, investing in paired spectrum seems to be a relatively save option.