In the thesis, we investigate the possible time-varying
structure of DSGE models currently used for monetary policy
analysis and forecasting. In order to replicate the observed
data, models are often equipped with additional exogenous
processes (technologies). These sector technologies are aimed
to capture some sector-specific (and often time-varying) part
of an economy’s behaviour. In the thesis, we extend a
relatively rich small open economy model with a set of
technologies which are tailored directly to the Czech data. We
find that the movement of technologies is a reflection of
variability of structural parameters and thus technologies'
incorporation enables us to keep structural parameters
relatively stable in time. Hence, such models can be regularly
used for policy analyses and forecasting without having to work
explicitly with time-varying structural parameters and
nonlinear filtering.