Estimation of the Czech Macroeconomic Model by Bootstrap Filter

J. Trinkewitz, J. Štecha, V. Havlena, O. Vašíček, and H. Pytelová (Czech Republic)


Bootstrap filter, Bayesian state estimation, Smoothing, Monte Carlo methods, macroeconomic model, monetary policy.


The monetary policy problem is explained in a simple the oretical framework of economy. The article presents a the oretical macroeconomic model and it shows the theoreti cal procedure of finding the optimal monetary policy un der discretion. The model is quantitatively analyzed on the data of the Czech economy. Model parameters are esti mated simultaneously by the original "extended bootstrap filter smoother". Monte Carlo methods can be used as a computational engine for a Bayesian state estimation for non-linear and/or non-gaussian systems. Using large number of samples, rep resentation of prior and posterior conditional probability function are obtained. Weighted bootstrap filter is a gen eral method for updating the samples representing the state estimate for a discrete time system. The contribution of this article lies in utilization of this procedure for state and parameter estimation of the macroeconomic model of the Czech republic. Weighted bootstrap algorithm is extended to the problem of smooth ing and thus a state estimation based on the whole set of measured data is obtained. It is particulary important in the case of the Czech macroeconomic model estimation, be cause only limited set of data is at disposal.

Important Links:

Go Back