A0992
Title: Stripping the discount curve: A robust machine learning approach
Authors: Markus Pelger - Stanford University (United States) [presenting]
Damir Filipovic - EPFL and Swiss Finance Institute (Switzerland)
Ye Ye - Stanford (United States)
Abstract: A robust, flexible and easy-to-implement method is introduced for estimating the yield curve from Treasury securities. The non-parametric method learns the discount curve in a function space that we motivate by economic principles. An extensive empirical study on U.S. Treasury securities shows that the method strongly dominates all parametric and non-parametric benchmarks. It achieves substantially smaller out-of-sample yield and pricing errors while being robust to outliers and data selection choices. The superior performance is attributed to the optimal trade-off between flexibility and smoothness, which positions our method as the new standard for yield curve estimation.
