Abstract Making investment decisions in general is a decision-making problem under uncertainty. How well an actual investment portfolio performs depends on the future evolution of economic and financial variables such as interest rates, asset returns and inflation rates. The future evolution of these risk drivers is traditionally modelled using time series models, and it is assumed that historical data are relevant for assessing future risk and return. However, opinions vary about the extent to which all forward-looking information can be derived from historical data. Consequently, a framework for combining views and model-based (density) forecasts is indispensable. We present a concrete example of how views can consistently be combined with model-based (density) forecasts and how this affects investment decisions.