A medium-scale nonlinear DSGE model is estimated (54 variables, 29 state variables, 7 observed variables). The model includes stock market. RMSE of in sample and out of sample forecasts are calculated. The nonlinear DSGE model with measurement errors outperform AR(1), VAR(1), linearized DSGE in...

Recent research shows that time-varying volatility plays a crucial role in nonlinear modeling. Contributing to this literature, we suggest a DSGE-GARCH approach that allows for straightforward computation of DSGE models with time-varying volatility, where the volatility component is formulated...

New simple forms of deviation from rational expectations (RE) are suggested: temporary near-rational expectations (TNRE) and persistent near-rational expectations (PNRE). The medium-scale DSGE model was estimated with the RE, the TNRE and the PNRE. It was estimated with and without observations...

This article suggests and compares the properties of some nonlinear Markov-switching filters. Two of them are sigma point filters: the Markov switching central difference Kalman filter (MSCDKF) and MSCDKFA. Two of them are Gaussian assumed filters: Markov switching quadratic Kalman filter...

In price formation it has become customary to consider two main types of price stickiness: by Rotemberg (Rotemberg (1982)) and by Calvo (Calvo (1983)). The nonlinear DSGE models (with different types of stickiness) are estimated for 11 different countries (both developed and emerging markets)....