Modeling with Impact

Macroeconometric Modeling, Forecasting, and Policy Analysis Using EViews

 

Monday

What will you learn and implement?

Model, as a tool to answer certain questions

History and generations of macroeconometric modeling

Theoretical framework and stylized facts of macroeconometric modeling

Types of macroeconometric models (structural, data-driven, hybrid, long-term vs short-term, quarterly vs annual, etc.)

Structure of macroeconometric models -blocks and their interactions

Types of data

Time series data. Nature and frequency

Data collection and processing/transformations

Working with Data in EViews

How to set up an EViews Work file.

How to input data to EViews

How to conduct a graphical analysis in EViews

Descriptive statistics and tests in EViews

Tuesday

What will you learn and implement?

Least squares estimation in EViews

The assumptions of least squares and hypothesis testing
How to conduct testing in EViews:

Normality test

Serial correlation test

Heteroskedasticity test

Functional form test

Stability test

To test linear restrictions.

Violating the Assumptions of Least squares

Autocorrelation

Heteroskedasticity

Multicollinearity

Misspecification

Correct the standard errors in case of heteroskedasticity and/or serial correlation.
Consequences of omitted or redundant variables and how to test them in EViews?
How to detect and account for outliers.

Wednesday

What will you learn and implement?

Nonstationarity and Unit root

Unit root tests in EViews

Making nonstationary variable stationary by differencing

Order of integration

De-trending trend stationary variables to make them stationary.

Why is testing for cointegration important: long-run estimation versus spurious regression. Error correction

Residual-based test for cointegration.

How to estimate FMOLS, DOLS, and CCR in EViews

How to conduct the Engle-Granger test in EViews

Single equation-based test for cointegration

How to estimate Auto Regressive Distributed Lags (ARDL) model in EViews

How to conduct the ARDL Bounds test in EViews

System of equations-based test for cointegration

How to estimate Vector Autoregressive (VAR) and Vector Error Correction (VEC) models in EViews

How to conduct the Johansen test in EViews

Error correction modeling and short-run estimation

General to specific (or David Hendry) modeling approach.

Brief information about Autometrics - a machine learning modeling algorithm

Thursday

What will you learn and implement?

Estimating the behavioral equations of macroeconometric model:

Theory-driven approach - structural models

Data-driven approach - statistical models

Combined approach - hybrid models

Corrections for outliers and structural breaks

Building a protype macroeconometric model in EViews:

Estimation and testing for behavioral equations: private consumption, government expenditure, and private investments

Putting behavioral equations and identities together to make a model.

The algorithms for solving models: Newton versus Gauss-Seidel versus Broyden.

Types for simulation and solutions for dynamics

Checking the validity and consistency of the completed model.

Checking statistical coherence of the model through statistical tests

Evaluating the predictive ability of the model: in-sample

Evaluating the predictive ability of the model: out-of-sample forecasting/projection

Calibration - making adjustments to the model to improve it. Add factors.

Updating the model: data and equations.

Friday

What will you learn and implement?

Understanding inputs and outputs of the model in EViews

Exogenous and endogenous variables, their switch

Designing scenarios - how to create scenario inputs in EViews.

How to solve the model for a given scenario using different solve options.

Making what-if simulations using a policy or exogenous variable to assess the effect of a policy

Making iterative simulations using a policy or exogenous variable to rich a policy target set

Options for reporting the simulation results: level versus growth rate in deviations.

Interpretations and policy implications