Model-Based Monetary Policy Analysis and Forecasting (MPAFx)

Source details

Session No.: OL 24.123
Delivery Method: Online Training
Location: Course conducted online
Date: January 1, 2024 - April 15, 2024 (15 weeks)
Primary Language: English
Participants are expected to have a background in undergraduate macroeconomics, statistics, and econometrics.
Participants are provided with guidance on how to access Matlab or Octave software.

All government officials are welcome to register.
The course is particularly well-suited for officials in central banks that are in the early stages of introducing Forecasting and Policy Analysis Systems (FPAS) with IMF assistance.
The course is offered in English.

Presents a semi-structural macroeconomic model often used as a core of FPAS at central banks.
Demonstrates implementation of the key equations of a canonical quarterly projection model (QPM) in macroeconomic modeling software.
Uses detailed country data highlighting an inflation targeting central bank for hands-on historical filtration, forecasting, and calibration exercises.
Covers two main technical aspects:
- Introduction to a canonical New Keynesian model structure and its key properties.
- Implementation of the QPM in Matlab/Octave and application of IRIS toolbox for solving and maintaining the QPM, as well as forecasting with the QPM.

Explain the key building blocks of a canonical semi-structural QPM.
Interpret the key model equations from a macroeconomic point of view.
Implement a simple QPM using a specialized software for macroeconomic modelling.
Distinguish the key elements of a QPM in a state-space form (i.e., shocks, observable and unobservable variables, measurement and transition equations, steady-state parameters, equation coefficients).
Identify necessary codes for data transformation, filtration and evaluation of the QPM properties.
Apply the basic IRIS Toolbox functions for solving the model.
Create output reports using model codes.
Develop a basic calibration of the QPM.
Produce a baseline forecast and alternative scenarios using the QPM.