SMM466 Applied Empirical Accounting 2021
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
MSc International Accounting & Finance
SMM466
Applied Empirical Accounting
2021
Answer FOUR questions.
ALL questions carry equal marks (25 marks)
The file data_exam.xlsx contains the series of daily prices for the S&P500 market index (SP), Chevron (CVX), West Texas Intermediate crude oil (WTI), Brent Dated (BRENT), covering the period 08/11/1990 – 2/11/2020.
QUESTION 1 (Unit Root Tests)
Import the dataset in Stata. Create a new dataset containing quarterly averages of the original series. Then, in the new dataset, generate four new series: spq for the natural log of quarterly S&P500, cvxq for the natural log of quarterly Chevron, wtiq for the natural log of quarterly WTI, and brentq for the natural log of quarterly
Brent.
DESCRIBE formally the testing procedure to determine the order of integration of the log prices series. PERFORM the relevant test for each log price series using Stata, report the tables of results and comment on the results.
(25 marks)
QUESTION 2 (Cointegration vs Spurious Regression)
Import the dataset in Stata. Create a new dataset containing quarterly averages of the original series. Then, in the new dataset, generate four new series: spq for the natural log of quarterly S&P500, cvxq for the natural log of quarterly Chevron, wtiq for the natural log of quarterly WTI, and brentq for the natural log of quarterly Brent.
Do wtiq (dependent variable) and brentq (independent variable) cointegrate? Or rather, do they spuriously correlate? IMPLEMENT the 2-step Engle and Granger procedure using Stata. REPORT and COMMENT on the results. EXPLAIN the difference between cointegrated versus spurious relationship.
(25 marks)
QUESTION 3 (Empirical Regression)
Import the dataset in Stata. Create a new dataset containing quarterly averages of the original series. Then, in the new dataset, generate four new series: spq for the natural log of quarterly S&P500, cvxq for the natural log of quarterly Chevron, wtiq for the natural log of quarterly WTI, and brentq for the natural log of quarterly Brent. After that, generate the returns series as follows: rspq=100*D.spq, rcvxq=100*D.cvxq, rwtiq=100*D.wtiq and rbrentq=100*D.brentq.
Estimate the following model:
Tcvxqt = F0 +F1Tspqt +F2Twtiqt +ct
COMMENT on the results of the estimation, considering the significance of the individual estimated coefficients and the general goodness of fit of the model. TEST at the 5% significance level the null hypothesis H0:F1 = 1 against the alternative H1:F1 想 1 using the appropriate test procedure.
PERFORM the Breusch-Pagan (1979) / Cook-Weisberg (1983) test and the Breusch-Godfrey (1978) test. Formally DESCRIBE the two tests and COMMENT on the results obtained.
(25 marks)
QUESTION 4 (ARMA Models)
Import the dataset in Stata. Create a new dataset containing weekly averages of the original series. Then, in the new dataset, generate four new series: spw for the natural log of weekly S&P500, cvxw for the natural log of weekly Chevron, wtiw for the natural log of weekly WTI, and brentw for the natural log of weekly Brent. After that, generate the returns series as follows: rspw=100*D.spw,
rcvxw=100*D.cvxw, rwtiw=100*D.wtiw and rbrentw=100*D.brentw.
Consider the series rspw and rbrentw. Using the Box and Jenkins (1970) approach find for each series the ARMA(p,q) model that fits the data “best” . Describe each step of the Box and Jenkins (1970) approach and report all the relevant plots and the tables of results.
(25 marks)
QUESTION 5 (Empirical Regression)
Import the dataset in Stata. Create a new dataset containing weekly averages of the original series. Then, in the new dataset, generate four new series: spw for the natural log of weekly S&P500, cvxw for the natural log of weekly Chevron, wtiw for the natural log of weekly WTI, and brentw for the natural log of weekly Brent. After that, generate the returns series as follows: rspw=100*D.spw,
rcvxw=100*D.cvxw, rwtiw=100*D.wtiw and rbrentw=100*D.brentw.
Estimate the following model
Twtiwt = F0 +F1Tspwt +F2TbTentwt +et
COMMENT on the results of the estimation, considering the significance of the individual estimated coefficients. DESCRIBE the Regression F-test and comment on the results reported in the Stata regression output. Predict the series of residuals and check Normality by creating the histogram and by performing the relevant Normality test. PERFORM the RESET test. Formally DESCRIBE the RESET test and COMMENT on the result obtained.
(25 marks)
2022-08-02