I can comment generally on it and say that what the numbers were based on was historical accident data. I can say that there have not been a great number of rail accidents with significant damages. So the data is limited to some extent, and obviously that's a positive thing, but it's based on the cost of actual accidents that have occurred and what the payouts were for them. That's how we came to the billion dollars.
I guess I'd just make another comment with respect to that. These are minimum requirements, so if class 1 railways choose to ensure their operations for more than a billion dollars, they are certainly free to do that. In terms of what their insurance levels are now or what the market supports, those things vary somewhat over time, but the billion dollars was arrived at by looking at historical accident data and what has actually been paid out, and what we understand this billion dollars would cover.