It's based, as I said, on a backcasting. It's looking at three or four years of transactions and then playing out with an economic model what the drop-off would be in the number reviewed had the billion-dollar threshold been in place in the past.
On May 21st, 2013. See this statement in context.