Interest rate term structure modeling using free-knot splines

Abstract

In this article a new methodology for estimating the term structure of interest rates is developed. Using polynomial splines, a reliable approximation to term structure may depend crucially upon intelligent selection of numbers and position of spline knots, which can be a combinatorially very complex task. A different approach based on heuristic optimization techniques called genetic algorithms is presented. The optimal spline function takes into account the goodness of fit of the spline function. The new methodology was applied to estimating the term structure using data on zero-coupon Euro market bonds.