top of page

Analysing inflation with semi-structural models

This chapter, prepared for the Research Handbook of Inflation, discusses semi-structural time series models for the analysis, forecast and now-cast of inflation. We define a semi-structural time series model as a multivariate structural time series model in the tradition of Harvey (1985) and Harvey (1990), where minimal economic restrictions are used to identify common and idiosyncratic trend and cyclical components of the observable data. We discuss the potential of this approach for inflation conjunctural analysis, forecasting and nowcasting in comparison with more widely used models in empirical macroeconomics such as factor models and VARs.

Download PDF • 4.75MB

Download PDF • 190KB


bottom of page