Calibrating the credit model: targeting the spread distribution
Document ID: 2010-1828
Published on: 10th May 2010
Author: Aubrey Clayton and Nick Jessop
This note describes a new calibration method for the Barrie & Hibbert credit model which follows a numerical integration approach based on the cumulative distribution function of π. Our new method allows us to calibrate µ and σ to match targets of mean spread levels and spread volatilities at a specified time horizon (typically either one year for ‘conditional’ calibrations or 100 years for “unconditional” calibrations). Note that α remains unconstrained by this approach, so we set it to a fixed value calibrated to historical data.
