Forecasting inflation is of key relevance for central banks, not least because the objective of low and stable inflation is embodied in most central banks’ mandates and the monetary policy transmission mechanism is well known to be subject to long and variable lags. To our best knowledge, central banks around the world use conditional as well as unconditional forecasts for such purposes. Turning to unconditional forecasts, these can be derived on the basis of structural and non-structural models. Among the latter, vector autoregressive (VAR)-models are among the most popular tools.

This study aims at assessing and deriving a set of unconditional forecasts for euro area inflation based on several specifications which take into account the information content of, inter alia, monetary and credit variables. The models are ordered and based on their in-sample performance and the “best” model is selected accordingly. The results indicate that the inclusion of money and credit variables in the information set improves the quality of the forecasts over a horizon of one to eight quarters. This supports the view that central banks should regularly monitor developments in money and credit.