How large is the tax clientele bias in the term structure of UK interest rates?

Abstract

This paper briefly explains why there may be a 'tax clientele' induced bias in the observed yield curve and attempts to estimate the size of this tax clientele bias in the term structure of UK interest rates. A method is based on the econometric technique developed by Nelson and Siegel (Journal of Business, 60, 473-89, 1987) is used to estimate yield curves, which permits observation of the magnitude of tax clientele bias. Estimates based on this method are presented using bond price information on 100 market days. The main finding is that the tax clientele effect is empirically small and therefore may not bias judgements based on the slope of the yield curve.