2) Instructions:
a. A brief description of data collection and data preparation process
b. Estimation and forecasting performance comparison of several different models.
The set of models should include at least one not autoregressive model
c. Out-of-sample forecasts / scenario forecasts based on estimated models
d. Some discussion of results
e. A set of interconnected indicators should be forecasted
f. Thorough preliminary analysis of data (for the presence of seasonality, working with outliers and missing observations) with all the necessary plots
g. The set of models should include a VAR / BVAR model with discussion of its results (including impulse response functions)
h. Forecasting performance should be evaluated using cross-validation