The Yield Curve: A Methodological Review and New Approximations for Estimation

Abstract

The yield curve is a tool widely used by those who make monetary policy decisions or plan their investments according to valuation, negotiation, or coverage on financial instruments. Because of its importance, our interest is focused on evaluating the performance of a set of econometric models for adjustments in the term structure of interest rates (in the context of the public debt market bond rates in Colombia and United States) and on the possible forms that the yield curves can take. The results reveal the goodness-of-fit of the artificial neural networks (ANN), the Svensson curve, the Nelson-Siegel curve, and local polynomials. Nevertheless, we strongly recommended the use the Svensson curve in the estimation of interest rates, due to the interpretability of its parameters and its superiority over the Nelson-Siegel Curve.

Note: Downloadable document is in Spanish.

Keywords: yield curves, Nelson-Siegel, Svensson, Kernel regression, smooth splines, local polynomials, Friedmann supersmoother, trigonometric polynomials, artificial neuronal networks

JEL Classification: C29, C51, C45, C53