Three essays on the study of macroeconomic variables using time series models

This dissertation is about examining macroeconomic variables using time series models. The first essay focuses on comparing the in-sample and out-of-sample properties of linear and nonlinear Taylor rules using real-time U.S. data and selecting the best form of the Taylor rule for different sample periods. It is found that the form of the Taylor rule differs across the pre-Volcker and the Volcker-Greenspan periods, and that there is increased coefficient instability in the Taylor rule during the Volcker-Greenspan period.;The second essay concentrates on decomposing U.S. real GDP into the trend and the cyclical components while accounting for unknown structural changes in the trend. The flexible Fourier form is introduced into the trend function to allow for gradual effects of unknown numbers of structural breaks occurring at unknown dates. The results suggest that there are multiple breaks in the trend and that the trend is deterministic once the breaks are accounted for. The implied cycle corresponds very closely to the NBER chronology.;The third essay identifies two types of supply shocks and two types of demand shocks, one of which is a monetary policy shock, using a four-variable Vector Autoregression (VAR). Two alternative measures of monetary policy shocks are considered: the federal funds rate and a new exogenous measure. The results from the VAR using the federal funds rate as the measure of monetary shocks suggest endogeneity in this measure of monetary shocks. The results from the VAR with the new exogenous measure of monetary policy shocks suggest that supply shocks were the main source of fluctuations in the U.S. real economy.