There are essentially three approaches to macroeconomic forecasting: judgemental forecasting, model forecasting, and statistical forecasting. Judgemental forecasters gather various kinds of information and data from official sources, and then they forecast future macroeconomic activity based on their own informal judgement about how the economy works.

Both model forecasting and statistical forecasting are done using sophisticated macroeconomic models and econometric models. Since this is a 2nd year subject, I believe it is beyond the scope of the class to ask the students to do either model forecasting or statistical forecasting. So, judgemental forecasting would be good enough as long as you give a detailed explanation following each number that explains why you chose that number. For example, you can use leading indicators to predict where the economy is headed.

To summarize, the report should be an empirical (not econometric) analysis rooted in macroeconomic theory.

2. How to detrend the data?

You would need to detrend the data for most aggregate variables (such as GDP, consumption, investment, etc). The best method of detrending the data would be a Hodrick-Prescott filter. However, since most of you are not familiar with this method, a good alternative would be expressing these aggregate variables in growth terms. For instance, instead of plotting the time series of real GDP, you should first remove the trend by calculating the real GDP growth rate and then plot the time series of real GDP growth rate. To see how to calculate growth rates, please read the file How to calculate growth rate (growthrates.pdf) that has been posted under learning resources on Blackboard.

3. How to calculate correlations and cross correlations?

An example of how to use Microsoft Excel to calculate these statistics have been posted on Blackboard (example.xls). Please note that when calculating correlations (and cross correlations), you are only required to calculate the correlation (and cross-correlation) of a particular variable (for example, unemployment rate) with output.

Last Updated on February 10, 2019 by EssayPro