In the preceding analysis I discussed how different loan and borrower characteristics defined the nonagency sectors. For the most part, I focused on minimum acceptable levels and sector averages. However, it is actually the distribution of these various factors that is most important. I say this because most of these variables have an impact on credit performance and speed in nonlinear ways. For example, defaults may increase x% going from a FICO score of 700 to 650 but increase 3x% moving an equal 50 points from 650 to 600. Thus it is the distribution of variables and, in particular, the lower cutoff point that is as important as the overall average.
Was this article helpful?