Stripping the Discount Curve — a Robust Machine Learning Approach

Abstract

We introduce a robust, flexible and easy-to-implement method for estimating the yield curve from Treasury securities. Our non-parametric method learns the discount curve in a function space that we motivate by economic principles. We show in an extensive empirical study on U.S. Treasury securities, that our 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. We attribute the superior performance to the optimal trade-off between flexibility and smoothness, which positions our method as the new standard for yield curve estimation.

Keywords: yield curve estimation, U.S. Treasury securities, term structure of interest rates, non-parametric method, machine learning in finance, reproducing kernel Hilbert space

JEL Classification: C14, C38, C55, E43, G12