Cost benefit fail December 6, 2013 at 11:34 am

From the Guardian:

The bill for cleaning up the huge Sellafield nuclear plant in Cumbria will rise even higher than its current estimated level of £70bn as operators struggle to assess the full scale of the task, according to sources close to the project.

It is the season to be generous so I will merely point out that this was entirely predictable (and predicted), and that it serves as a useful lesson about the risks of making long term infrastructure decisions on the basis of cost benefit analysis where both sides are subject to oodles of model risk.

2 Responses to “Cost benefit fail”

  1. It is very reassuring that HS2 has learnt this lesson. More seriously, is there a credible alternative to CBA? Or is it a case of it being least bad but application needs to be better (a la regulatory capital and RWAs)?

  2. Well, at least you could honestly try to produce probability fan charts of the costs and benefits, and hence derive an implied equity (which has a much higher cost) to support the project risk. It isn’t perfect and there is still a lot of model risk but at least you are accepting that there is uncertainty in the outcome and trying to price it.