The ability to attribute portfolio risk and performance to key factors, such as overall market exposure, rates, sectors, and quality, is an essential tool for helping portfolio managers to understand their risk and interpret their results. A parsimonious factor risk model can also support advanced portfolio construction goals, such as minimizing benchmark tracking error or realizing factor exposure tilts. It is notoriously difficult, however, to build such models for bond portfolios, as a myriad of data quality concerns arises, driven by a vast, frequently illiquid market. Advanced modeling techniques are required to trim outliers and infer term structure shapes from limited and noisy data, so that the factor returns used to measure portfolio risk reliably capture systemic risk rather than noise.
The Axioma Factor-based Fixed Income Risk Model improves upon current risk modeling techniques in a number of innovative ways.