Okay.
I understand that, but to go back to my original point, it's still the case that in 2009-10 you ended up spending $1.2 billion less than you thought you would spend at the beginning of the year. I don't understand why it's so asymmetrical. If you base your estimates on when you think the equipment will be delivered, then some years it might come sooner than expected and other years it might come later, but in your case it seems to be always much later than you expect.
Why don't you adjust your expectations to match the reality and then you wouldn't have these huge lapses?