First we recognized lots of defaults were a systemic good
Now we need a model for defaults that can guide credit policy
A rough notion of a better path for the aggregate default Rate is necessary but inadequate
It's the optimal path and pattern of actual defaults we seek
