The European Central Bank (ECB) relies on the timely availability of high-quality financial market indicators to facilitate data-driven decision-making. Gauges of interest rate dynamics are of particular interest to the central bank, given their relevance for monetary policy and financial stability. By capturing the term structure of interest rates, zero-coupon yield curves serve as the key barometer of market expectations of monetary policy, economic growth, and inflation. Correspondingly, the prevailing practice among major central banks, including the ECB, is to maintain an internal system for estimating yield curves. This thesis aims to contribute to three areas relevant to implementing such systems. First, a natural generalization of the Nelson-Siegel model based on Laguerre polynomials is evaluated as an alternative to the popular Svensson extension. The evidence from the conducted evaluation study points to several theoretical and empirical arguments in favor of such an approach, particularly in the areas of model stability, flexibility, and out-of-sample forecasting performance. Second, the empirical study evaluates the nature of the trade-off between goodness-of-fit and model stability resulting from applying different methodologies to estimating the nonlinear decay parameter. Finally, a computationally efficient method based on iterated least-squares is considered for fitting the zero-coupon yield curve to a sample containing fixedcoupon bonds, under a fixed nonlinear decay parameter. The method provides an alternative to the common practice of minimizing either yield-to-maturity error or (weighted) price error.