September 26, 2017

SMOOTHIES: A Toolbox for the Exact Nonlinear and Non-Gaussian Kalman Smoother

In this paper, I present a new toolbox that implements the exact nonlinear and non-Gaussian Kalman smoother for a wide class of discrete-time state space models, including models with implicit functions and equality constraints.

Read also CPB Discussion Paper 359.

The technical details are presented in an accompanying paper, while the toolbox is documented in the current one. The toolbox, which is built on top of Dynare, is very user-friendly and only requires the user to provide the state space model to be analyzed, while the toolbox automatically solves the smoothing problem. The toolbox can also be applied for conditional forecasting, which is demonstrated on the basis of a nonlinear macroeconomic model with forward-looking variables.

Authors

Joris de Wind