Hello, if you have any need, please feel free to consult us, this is my wechat: wx91due

FIN4013 Financial Forecasting

2024/2025 Semester 2

Assignment 1 (Due: 5pm, Mar 12, 2025)

Q1. How can a seasonal data be forecast by a linear regression model? Explain with an example model.

Q2. The following regression results relate to a study of the salaries of public school teachers in a mid-western city (the sample size was 450 teachers):

R-squared = 0.684;
Standard error of the estimate = 2,000.
EXP is the experience of teachers in years of full-time teaching.

a. What is the t-ratio for EXP? Does it indicate that experience is a statistically significant determinant of salary if a 95 percent confidence level is desired?

b. What percentage of the variation in salary is explained by this model?

c. Determine the point forecast of a salary for a teacher with 20 years of experience.

d. What is the approximate 95 percent confidence interval for your point forecast from (c)?

Q3. Consider the simple time trend model  (t=1,2,…, T) that can be written with matrix notation as . Write out the matrix expression for  in terms of T. Hint: using the following results,

Q4. Modelling and Reporting

Refit the US GDP growth prediction models (in tutorial exercise 2 of chapter 1- Forecasting US GDP Growth) using the following dynamic specification

while the explanatory variables, estimation and forecast sample periods remain the same as those in the tutorial exercise 2 of chapter 1. Answer the following questions.

(a) Explain how you revise the code for the dynamic specification?

(b) Run your revised codes and report the estimation sample measures of fit of the dynamic models as in the example of Table 1.13.1 in textbook (or in slide 72 of chapter 1 PPTs);

(c) Run your new codes and report the forecast evaluation statistics of static and recursive forecasts of the dynamic models as in the example of Table 1.13.2 in textbook (or in slide 73 of chapter 1 PPTs);

(d) Have a brief result discussion on the three dynamic models with different explanatory variables.

(e) Comparing to the baseline linear regression model in tutorial exercise 2 of chapter 1, do you think the dynamic specification is superior in prediction? Explain why?