Mr. Alex Chisholm:

We were quite careful not to offer any forecast. Instead we set some minimum prices in accordance with an internationally benchmarked approach. Our advisers provided us with a database of all the results ever obtained in a spectrum auction, to which we made various adjustments for the size of the Irish economy, the amount of spectrum available, the number of firms in the market and so on. From these data we came up with a conservative minimum estimate of what the payments were likely to be for each part of the spectrum that was on sale. That was how we priced it.

What we could not know was how much of an excess there might be. That depended on the firms and their private, never revealed to us valuations of the spectrum, their business strategies and also how many bidders there would be. We could see, for example, that the 800 MHz band, which is the new spectrum that was freed up by the move from analogue services to digital terrestrial broadcasting, was attracting a great deal of interest. There was 30 MHz of paired spectrum available within that band in blocks up to 5 MHz or 10 MHz, with a limit of 10 MHz for each bidder. This meant that if there were three bidders, everything would fit quite neatly, but less so if there were four. Factors like that make a huge difference to the competitive edge.

As it turned out, our minimum estimate was comfortably exceeded. The lesson there is that no matter how much one consults and dialogues with firms, one will not discover their private valuation for the spectrum until a competitive auction is held. This is one of the strongest reasons for having an auction for scarce assets.