Expectations measured by surveys are often available only as fixed-event forecasts (e.g., forecasts for the end of the current and next calendar years). Yet, the interest may be to track the evolution of rolling-event forecasts (e.g., forecasts for the next 12 months). A common practice is to approximate the rolling-event forecasts as an average between two fixed-event forecasts, but despite its popularity the literature provides no formal justification for this procedure. This note offers such a justification and additional insights to map from fixed-event to rolling-event forecasts. The methods are illustrate with the computation of inflation expectations at various horizons of interest