Overview of structural modeling tools

Slang: model source language

The model source language is used to describe all the elements of a model in human-readable form: variables, parameters, shocks, equations, etc. The model source is then translated to a structural model object.

Structural models

Structural models are systems of dynamic simultaneous (interdependent) equations with lags and leads (expectations). Iris supports nonlinear nonstationary (balanced growth path) structural models.

Simulation plans

Simulation plans define more complex simulation assumptions for various types of models (structural models, explanatory equations, vector autoregressions): anticipation status, inversion pairs (exogenize/endogenize) and conditioning information.

Explanatory equations

Explanatory equations are systems of sequential (non-simultaneous) equations that can be estimated and executed (simulated) one after another. The advantage over the structural models is faster evaluation, especially for nonlinear equations.

State space systems

Linear time-varying state-space systems can be used for designing nontrivial Kalman fitering tasks.

Nonlinear equations solver settings

Iris features its own nonlinear equations solver used in calculating the steady state and dynamic simulations of structural models.