Thanks, this helps me understand. So you exclusively model bad outcomes that require restrictions and omit just-as-likely outcomes that would not require restrictions?

— Fraser Nelson (@FraserNelson) December 18, 2021