invested one year earlier (see Figure 9). Thus the value assigned to the payout variable in
the Citéfibre case is obtained by multiplying by three the foregone revenues suffered in the
period in which the decision to postpone was made.
Per unit up-front costs are estimated by resorting to a study of the investment costs of
NGN published the French telecommunications regulator Arcep (2006): the average cost
estimated by Arcep (€500) has been decreased to take into account that in Paris the roll-out
of the fibre is less expensive due to the possibility of exploiting the sewage system. A
sanity check of this value has been carried out by comparing with the value inferred
directly using Citéfibre available data. The up-front cost is therefore assumed to be equal to
€400.
Figure 9 - Revenues lost by investing one year later
The annualized volatility of the log return has been estimated using bi-weekly stock
returns. Since Citéfibre was floated on Euronext on the 2nd December 2005, the sample size
is slightly above 45 observations. A higher frequency would have increased the accuracy of
the estimate by increasing the number of observations; however this might have entailed an
24 Prior to December 2005 Citéfibre was not traded and thus this is the first market capitalisation availble.
26