代写计量经济的bonus小作业Tools:eviews题目：Term Empirical Exercise (Optional/Bonus)
Introduction to Econometrics, 2017/2018-1
Term Empirical Exercise (Optional/Bonus)
Number of students: 1 to 3
Due date: December 15, 2017, 17:00
Documents required: Report & EViews workfile
1. Identify a financial issue that you would like to investigate with reference to the literature (academic journal papers). Take note about the variables involved, and make sure that you can collect both cross-sectional and time series data for this issue.
2. Choose a representative academic journal paper in the literature of your issue, and take note of the “points to consider when reading an empirical paper”.
3. Set up the econometric model, taking note of the a priori expectation of the parameters and the assumptions of the models. Set up different models, at least one bivariate regression model and one multiple regression model. You may also use models with variables in different functional forms.
4. Carry out the least squares estimation of the regression models using both the cross-sectional data and time-series data, report the estimation results, and interpret the results. Are your parameter estimates correctly signed? What is the goodness of fit?
5. Carry out hypothesis testing in relation to the issue under study.
6. Do the assumptions hold? Are the error terms normally distributed? Do you suspect high or low degree of multicollinearity? Is there any heteroskedasticity problem in your cross-sectional data? Is there any autocorrelation problem in your time-series data?
7. How do you correct for the problem(s) that you have found in 5? Produce additional empirical results for the issue under study after correcting for the problem(s)? Compare the additional results with the previous results.
8. What is your conclusion on the issue under study based on your estimation and hypothesis testing results?
1. Identify a financial issue that you would like to investigate with reference to the literature (academic journal papers). Take note about the variables involved, and make sure that you can collect both cross-sectional and time series data