Stochastic simulation and dynamic properties of the new version of the Italian model

This paper describes the results of some stochastic simulation experiments performed on the most updated version of the Italian model. Due to a change in the income accounts system, the model has been completely reestimated using the new quarterly data. It consists of 128 equations, 50 of which are stochastic. As regards to the structure of the model, the main differences with respect to the previous version lie in the income sector: now the different components of income distribution are determined endogenously and disaggregated by sector, and affect directly private disposable income. Stochastic simulation has been performed using the program described in [1]. The generation of pseudo randorn numbers with multivariate normal distribution has been performed using the Box-Muller technique [3] and McCarthy algorithm [9].