Other stats such as how many addresses have move in the 6 to 2 months would help. I think It can be nailed with a reasonable margin of error