Econometrics Postgraduate

Choose a stock market index and unemployment rate in a given country. You can use data sources from Datastream, Google Finance, Federal Reserve Bank of St. Louis or others.

 

Use the data on your chosen market index and unemployment rate to study the relationship between these two variables. To do so, you need to follow these steps:

 

(i) [10 marks] Report and briefly analyse the summary statistics which describe the statistical properties of your series (in levels and logs). You also need to plot the series and describe patterns that you observe. For example, do you observe trends? If so, what type of trends?

 

(ii) [10 marks] Report and comment on the ACF/PACF functions of stock market index and unemployment rate and their first-differences in log.

 

(iii) [20 marks] Use appropriate Dickey–Fuller tests to test if the two variables stock market index and unemployment rate and their first-differences in log are non-stationary (have a unit root) or not. Explain the test procedure, report the test statistics and the appropriate critical values (at 5% significance level) and discuss the results.

 

(iv) [20 marks] Use a Johansen’s trace test statistic to test for cointegration between the two variables. Explain the test procedure, report the test statistic and the appropriate critical value (at 5% significance level). Discuss the results.

 

(v) [30 marks] Thereafter, if you have found a cointegrating relation, estimate the long-run causal relationship (i.e., the cointegrating relation) and the Error Correction Model. Otherwise, use the Box-Jenkins procedure to estimate a VAR model for the appropriately transformed (de-trended or differenced) variables. Report the results of Granger causality. Discuss the results.

 

(vi) [10 marks] Find two published papers on the relationship between stock market index and unemployment rate and discuss/compare your results to those obtained in these papers.

 

Overall word limit 1250 words maximum