The paper aims at finding the most optimal individual, collective, and combined yield curve forecasting models. It is shown that incorporating macroeconomic information improves the model's goodness-of-fit characteristics. It is also proved that combined forecasts perform better on average when...