We used the predicted range of consumption and production numbers for 2015 to give us the low case, middle case and high case projected net export decline rate from 2005 to 2015.

Once production in a given exporting country starts falling, the actual net export decline rate tends to accelerate with time, which is what we are almost certainly going to see for Saudi Arabia in 2007 versus 2006.

Sorry, I meant how did you get the +4.4% consumption number for Saudi Arabia? Is that the average of the last 2 years demand? Or three years? Or some other procedure? I don't doubt the results, I would just like to be able to explain how it was calculated if someone was to say, why shouldn't it be 2.2%.

Khebab used a Monte Carlo simulation method, based on recent consumption, to estimate a range of future consumption, which resulted in very low projections for Russia.