Business Cycle Models and Stylized Facts in Germany

The aim of this paper is to test to what extent a benchmark real and monetary business cycle model can account for some basic stylized facts with a particular emphasis on monetary variables. We calibrate the model on German data using the method proposed by Cooley and Prescott (1995). First we will analyze the dynamic properties of the models, the Impulse Response Functions and propose a variance decomposition (for the monetary BC Models). We find that even though money is not neutral in the short run, the effect of a monetary shock is only marginal compared to the productivity shock, i.e. the share of the variance of the monetary shock in the total variance of the forecast error is small and decreases rapidly. We simulate the models and compare the properties of the model economies with those of the observed data. The evidence suggests that the benchmark RBC model can account for some stylized facts in Germany. The general pattern of the relative volatilities of investment, output and consumption is replicated by the model. Nevertheless, the overall volatility is too high and the level of the relative volatilities is not well reproduced. The introduction of exogenous monetary shocks and a cash-in-advance constraint increases the relative volatilities and the cross correlation of consumption. In general the second order moments of money (M1) and inflation are not well reproduced.