FIN5201 FINANCIAL ECONOMETRICS 2021
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
FIN5201
FINANCIAL ECONOMETRICS
Question 1
Khairul made a statistical analysis to examine the impact ofthe economic freedom dimensions to the cost efficiency and profit efficiency on the X bank from 1990 to 2018 using the regression under the Ordinary Least Square (OLS) estimation method. The regression models on cost efficiency and profit efficiency are developed as below:
Cost Efficiency Model (Equation 1; M1)
CE = α0 + α1(BusFree)t + α2(LabFree) + α3(FinFree)t + α4(InvFree)t + α5(TaxBur)t + α6(GovtSpen)t + α7(GNI)t + α8(Inf)t + ut
Profit Efficiency Model (Equation 1; M2)
PE = α0 + α1(ProRigh)t + α2(GovtInt) + α3(JudEff)t + α4(FisHeal)t + α5(TaxBur)t + α6(GovtSpen)t + α7(GNI)t + α8(Inf)t + ut
a. Should Khairul consider the issue of autocorrelation? Explain
(4 marks)
b. The Breusch-Godfrey Serial Correlation LM test is describe as below,
Test result on Cost Efficiency:
Breusch-Godfrey Serial Correlation LM Test:
F-statistic 3.737761 Prob. F(8,108) 0.7644
Obs*R-squared 7.877864 Prob. Chi-Square(2) 0.5980
Test result on Profit Efficiency:
Breusch-Godfrey Serial Correlation LM Test:
F-statistic Obs*R-squared
1.445672 Prob. F(8,108) 0.0337
4.453071 Prob. Chi-Square(2) 0.0099
i. Does the issue of autocorrelation exist in cost efficiency and profit efficiency
regression analysis? Explain
(8 marks)
ii. If yes, how to solve this issue? Explain
(4 marks)
c. The results of Variance Inflation Factor (VIF) test for independent variables in both the models of cost efficiency and profit efficiency are shown below:
Variance Inflation Factors Test on Cost Efficiency Model:
Coefficient Variable Variance |
C BusFree LabFree FinFree InvFree TaxBur GovtSpen GNI Inf |
Variance Inflation Factors Test on Profit Efficiency Model:
Coefficient Variable Variance |
C ProRigh GovtInt JudEff FisHeal TaxBur GovtSpen GNI Inf |
i. Does the issue of multicollinearity exist in cost efficiency and profit efficiency models (assuming VIF value 5)?
(4 marks)
ii. Identify and explain which of the variables are having the multicollinearity issues in both models
(8 marks)
iii. Several approaches can be used to minimize or solve this issue including drop one of the collinear variables, transform the highly correlated variables into a ratio, increase the data sample, longer run of data or switch to a higher frequency. Explain another approach which could be practiced to solve this issue by keeping those variables and maintaining the observation sample
(4 marks)
d. Assuming equation model of Gross National Income model (GNI) (Equation 2):
GNI = β0 + β1(export)t + β2(employ)t + β3(CE)t + u2t
i. Develop the simultaneous Equation Model (SEM) in cost efficiency by inserting the Equation 2 into Equation 1; M1
(10 marks)
ii. Identify which of the variables are categorizes under exogenous variable. Explain.
(4 marks)
iii. Explain which ofthe variables has endogenity issue.
(4 marks) (Total 50 Marks)
Question 2
You are the expert analyst in the banking sector. The ABC Bank had appointed you to examine the potential determinants that may significantly influence the bank’s profitability. The following regression results based on a sample of212 branches of ABC Bank:
Yi
Std error
prob
= -2.761 + 0.312Xaajt + 0.091Xbjt - 0.153Xcjt R2 = 0.9881
= (0.050) (0.131) (0.056)
= (0.000) (2.341) (0.041)
where Y = bank’s profitability in Ringgit Malaysia
Xa = size of bank
Xb = leverage
Xc = bank risk
j = the individual bank branch
t = time period
a. Discuss in details for each variable in the regression result.
(9 marks)
b. Explain in specific on the result of R2 and any possible factors that contribute to the R2 in the estimation model
(10 marks)
c. At = 1%, test the hypothesis that the size of bank has no effect on the profitability of bank versus that it has a positive effect.
(6 marks)
(Total 25marks)
Question 3
The score is the dependent variable representing the average final exam scores ofthe postgraduate students. The estimation of linear probability model is shown below:
̂ = 0.321 + 0.543 log(income) + 0.772 log(experience) + 0.123 education
(0.755)
[0.754]
+ 0.022 age +
(0.057)
[0.057]
n = 608
(0.034)
[0.031]
0.0005 age2 (0.035) [0.034]
(0.005)
[0.005]
- 0.504 busyness (0.097) [0.096]
(0.021)
[0.020]
+ 0.734 gender
(0.006)
[0.006]
R2 = 0.078
The usual standard errors are reported in parentheses ( ) and the heteroskedasticity-robust standard errors are reported in brackets [ ].
a) The estimation model contains both sets of usual standard errors and heteroskedasticity- robust standard errors. Please discuss if there are any important differences between the two sets of standard errors
(4 marks)
b) By assuming ceteris paribus, discuss the impact to the score if:
i. the education decreases by five and half years
ii. the busyness increases by three years
(4 marks)
c) Explain, at what points does another year of age reduce the probability of score?
(4 marks)
d) By assuming ceteris paribus, interpret the coefficient of the gender’s dummy or binary variable (a dummy variable equal to one if the student is male and zero is female) and
explain how you relate it with the effect of having five and half years of education (4 marks)
e) The student number of 302 in the data set has the following characteristics: income = 7,500.87, experience =14.75, education = 13, age = 38.5, busyness =4 and gender = 0. Compute the expected scoring for this student result.
(9 marks) (Total 25marks)
2022-02-07