Predictive Density Combination Using a Tree-Based Synthesis Function Staff working paper 2023-61 Tony Chernis, Niko Hauzenberger, Florian Huber, Gary Koop, James Mitchell This paper studies non-parametric combinations of density forecasts. We introduce a regression tree-based approach that allows combination weights to vary on the features of the densities, time-trends or economic indicators. In two empirical applications, we show the benefits of this approach in terms of improved forecast accuracy and interpretability. Content Type(s): Staff research, Staff working papers JEL Code(s): C, C1, C11, C3, C32, C5, C53 Research Theme(s): Models and tools, Econometric, statistical and computational methods, Monetary policy, Real economy and forecasting