EMIANISHA BT MD SUMAN

Assignment 2 and 3 Financial Econometrics (20%)

U2102965

EII3005
Assignment 2 (Eviews) 10 marks [data set is attached]
1. Plot the three yield curve of the US Treasury Bill of 4-week, 3-month, and 6-month
USTB.
ustb3m
6
5
4
3
2
1
0
01 02 03 04 05 06 07 08 09

ustb4w
6

-2
01 02 03 04 05 06 07 08 09

ustb6m
6
5
4
3
2
1
0
01 02 03 04 05 06 07 08 09

1
Assignment 2 and 3 Financial Econometrics (20%)

2. Run unit root test ADF and KPSS test on ustb4w.

Augmented Dickey-Fuller Unit Root Test on USTB4W.
Null Hypothesis: USTB4W has a unit root
Exogenous: Constant
Lag Length: 11 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -0.814348 0.8144

Test critical values: 1% level -3.433278
5% level -2.862720
10% level -2.567444

*MacKinnon (1996) one-sided p-values.

H0: USTB4W has a unit root (non-stationary).

H1: USTB4W is stationary.
P-value = 0.8144 > a = 0.05, do not reject H0 at 5% significance level, USTB4W is non-
stationary series.
KPSS Unit Root Test on USTB4W.
Null Hypothesis: USTB4W is stationary
Exogenous: Constant
Bandwidth: 35 (Newey-West automatic) using Bartlett kernel

LM-Stat.

Kwiatkowski-Phillips-Schmidt-Shin test statistic 1.010135

Asymptotic critical values*: 1% level 0.739000
5% level 0.463000
10% level 0.347000

*Kwiatkowski-Phillips-Schmidt-Shin (1992, Table 1)

Residual variance (no correction) 2.522238

HAC corrected variance (Bartlett kernel) 88.68061

H0: USTB4W is stationary.

H1: USTB4W is non-stationary.
LM-statistics = 1.010135 > 0.739000 (1%), 0.463000 (5%), 0.347000 (10%), USTB4W is non-
stationary.

2
Assignment 2 and 3 Financial Econometrics (20%)

3. Take the 1st difference on ustb4w. Plot the graph.

Null Hypothesis: D(USTB4W) has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 10 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -16.40822 0.0000

Test critical values: 1% level -3.962434
5% level -3.411957
10% level -3.127881

*MacKinnon (1996) one-sided p-values.

H0: USTB4W has a unit root (non-stationary).

H1: USTB4W has no unit root (stationary).
P-value = 0.0000 < a = 0.05, reject H0 at 5% significance level, USTB4W is stationary after
taking the first difference.
D(USTB4W)
1.2

0.8

0.4

0.0

-0.4

-0.8

-1.2
01 02 03 04 05 06 07 08 09

3
Assignment 2 and 3 Financial Econometrics (20%)

4. Assuming all the three series are integrated of order one I (1), what kind of analysis can
you do?
A series with a unit root (a random walk) is said to be integrated of one order or I (1).
If all the series are I (1), we can use the Vector Error Correction Model (VECM).

5. Run the Johansen-Juselius Cointegration Test and show the result.

What is your conclusion?
Date: 12/17/23 Time: 22:41
Sample (adjusted): 8/09/2001 12/31/2009
Included observations: 2097 after adjustments
Trend assumption: Linear deterministic trend
Series: USTB3M USTB4W USTB6M
Lags interval (in first differences): 1 to 6

Unrestricted Cointegration Rank Test (Trace)

Hypothesized Trace 0.05

No. of CE(s) Eigenvalue Statistic Critical Value Prob.**

None * 0.049408 132.8667 29.79707 0.0000

At most 1 * 0.012562 26.61149 15.49471 0.0007
At most 2 4.82E-05 0.101143 3.841465 0.7505

Trace test indicates 2 cointegrating eqn(s) at the 0.05 level

* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

Unrestricted Cointegration Rank Test (Maximum Eigenvalue)

Hypothesized Max-Eigen 0.05

No. of CE(s) Eigenvalue Statistic Critical Value Prob.**

None * 0.049408 106.2552 21.13162 0.0000

At most 1 * 0.012562 26.51035 14.26460 0.0004
At most 2 4.82E-05 0.101143 3.841465 0.7505

Max-eigenvalue test indicates 2 cointegrating eqn(s) at the 0.05 level

* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

H0: r=0
H1: 0 < r ≤ g
Based on the p-value of the Trace statistic and Maximum Eigenvalue, having p-values that are
less than 0.05, the null hypothesis is rejected at a 5% significance level.
R=2, reject H0, there are 2 cointegrating vectors. We can proceed to the Error Correction Model.

4
Assignment 2 and 3 Financial Econometrics (20%)

6. Conduct the relevant analysis to check whether there is any information flow between the
three yield curves.
By inspecting the graph of USTB3M, USTB 4W, and USTB 6M, we can see that all three
variables are moving together. We can run Engle-Granger two steps approach to see if they
have a long-run relationship.
6

-1
01 02 03 04 05 06 07 08 09

ustb3m ustb4w ustb6m

To determine whether they have a long-run relationship, all the variables must be cointegrated of
the same order. After conducting the stationarity test, we found that all the variables become
stationary after taking the first difference. This means that all the variables are integrated of the
same order, I (1).

ADF: USTB3M (level).

Null Hypothesis: USTB3M has a unit root
Exogenous: Constant
Lag Length: 4 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -0.531500 0.8826

Test critical values: 1% level -3.433267
5% level -2.862715
10% level -2.567442

H0: USTB3M has a unit root.

H1: USTB3M is stationary.
P-value=0.8826 > a=0.05, do not reject H0, USTB3M is non-stationary.

5
Assignment 2 and 3 Financial Econometrics (20%)

1st difference.
Null Hypothesis: D(USTB3M) has a unit root
Exogenous: Constant
Lag Length: 3 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -24.09144 0.0000

Test critical values: 1% level -3.433267
5% level -2.862715
10% level -2.567442

H0: USTB3M has a unit root.

H1: USTB3M is stationary.
P-value=0.0000 < a=0.05, reject H0, USTB3M is stationary.
ADF: USTB4W (level).
Null Hypothesis: USTB4W has a unit root
Exogenous: Constant
Lag Length: 11 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -0.814348 0.8144

Test critical values: 1% level -3.433278
5% level -2.862720
10% level -2.567444

H0: USTB4W has a unit root.

H1: USTB4W is stationary.
P-value=0.8144 > a=0.05, do not reject H0, USTB4W is non-stationary.
1st difference.
Null Hypothesis: D(USTB4W) has a unit root
Exogenous: Constant
Lag Length: 10 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -16.40678 0.0000

Test critical values: 1% level -3.433278
5% level -2.862720
10% level -2.567444

H0: USTB4W has a unit root.

H1: USTB4W is stationary.
P-value=0.0000 < a=0.05, reject H0, USTB4W is stationary.
ADF: USTB6M (level).

6
Assignment 2 and 3 Financial Econometrics (20%)

Null Hypothesis: USTB6M has a unit root

Exogenous: Constant
Lag Length: 22 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -0.518868 0.8851

Test critical values: 1% level -3.433294
5% level -2.862727
10% level -2.567448

H0: USTB6M has a unit root.

H1: USTB6M is stationary.
P-value=0.88514 > a=0.05, do not reject H0, USTB6M is non-stationary.
1st difference.
Null Hypothesis: D(USTB6M) has a unit root
Exogenous: Constant
Lag Length: 21 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -8.257789 0.0000

Test critical values: 1% level -3.433294
5% level -2.862727
10% level -2.567448

H0: USTB6M has a unit root.

H1: USTB6M is stationary.
P-value=0.0000 < a=0.05, reject H0, USTB6M is stationary.
Estimate the long-run model.

❖ All three variables are integrated of order 1 or I (1).

7
Assignment 2 and 3 Financial Econometrics (20%)

Estimate the long-run model.

Dependent Variable: USTB3M
Method: Least Squares
Date: 12/17/23 Time: 21:50
Sample: 7/31/2001 12/31/2009
Included observations: 2104

Variable Coefficient Std. Error t-Statistic Prob.

USTB4W 0.456808 0.005481 83.33662 0.0000

USTB6M 0.551812 0.005566 99.14743 0.0000
C -0.058854 0.002769 -21.25181 0.0000

R-squared 0.998364 Mean dependent var 2.228165

Adjusted R-squared 0.998363 S.D. dependent var 1.585151
S.E. of regression 0.064137 Akaike info criterion -2.654164
Sum squared resid 8.642593 Schwarz criterion -2.646106
Log likelihood 2795.181 Hannan-Quinn criter. -2.651213
F-statistic 641241.5 Durbin-Watson stat 0.382424
Prob(F-statistic) 0.000000

1% increase in USTB4W will increase USTB3M by 0.46%.

1% increase in USTB6M will increase USTB3M by 0.55%.

Save the residual and test the stationarity of the residual.

Null Hypothesis: RESIDUAL_LONGTERM has a unit root

Exogenous: None
Lag Length: 10 (Automatic - based on SIC, maxlag=25)

t-Statistic Prob.*

Augmented Dickey-Fuller test statistic -7.990929 0.0000

Test critical values: 1% level -2.566067
5% level -1.940975
10% level -1.616598

H0: Residual from cointegrating regression has a unit root (non-stationary).

H1: Residual from cointegrating regression is stationary.

8
Assignment 2 and 3 Financial Econometrics (20%)

P-value=0.000 < a=0.05, reject H0, residual is stationary, therefore, there is a cointegrating
relationship between the variables.
VAR Lag Order Selection Criteria
Endogenous variables: USTB3M USTB4W USTB6M
Exogenous variables: C
Date: 12/17/23 Time: 22:37
Sample: 7/31/2001 12/31/2009
Included observations: 2096

Lag LogL LR FPE AIC SC HQ

0 -1236.426 NA 0.000655 1.182659 1.190742 1.185620

1 9771.982 21974.80 1.81e-08 -9.312960 -9.280625 -9.301116
2 9925.968 306.9424 1.58e-08 -9.451305 -9.394719 -9.430577
3 10004.36 156.0310 1.48e-08 -9.517516 -9.436680 -9.487905
4 10082.44 155.2051 1.38e-08 -9.583439 -9.478351 -9.544945
5 10125.80 86.05525 1.34e-08 -9.616224 -9.486885 -9.568846
6 10222.43 191.5088 1.23e-08 -9.699841 -9.546251* -9.643579*
7 10234.36 23.60915 1.23e-08 -9.702636 -9.524796 -9.637492
8 10251.48 33.82132* 1.22e-08* -9.710379* -9.508288 -9.636352

Two criteria suggest that lag 6 is an optimal lag and three criteria suggest that lag 8 as an optimal
lag. I choose lag 6.

Inverse Roots of AR Characteristic Polynomial

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5
-1 0 1
H0: The estimated VAR is stationary.
H1: The estimated VAR is non-stationary.
Since there is no inverse roots lie outside the unit circle, do not reject H0. Therefore, the estimated
VAR is stable.

9
Assignment 2 and 3 Financial Econometrics (20%)

Date: 12/17/23 Time: 22:41

Sample (adjusted): 8/09/2001 12/31/2009
Included observations: 2097 after adjustments
Trend assumption: Linear deterministic trend
Series: USTB3M USTB4W USTB6M
Lags interval (in first differences): 1 to 6

Unrestricted Cointegration Rank Test (Trace)

Hypothesized Trace 0.05

No. of CE(s) Eigenvalue Statistic Critical Value Prob.**

None * 0.049408 132.8667 29.79707 0.0000

At most 1 * 0.012562 26.61149 15.49471 0.0007
At most 2 4.82E-05 0.101143 3.841465 0.7505

Trace test indicates 2 cointegrating eqn(s) at the 0.05 level

* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

Unrestricted Cointegration Rank Test (Maximum Eigenvalue)

Hypothesized Max-Eigen 0.05

No. of CE(s) Eigenvalue Statistic Critical Value Prob.**

None * 0.049408 106.2552 21.13162 0.0000

At most 1 * 0.012562 26.51035 14.26460 0.0004
At most 2 4.82E-05 0.101143 3.841465 0.7505

Max-eigenvalue test indicates 2 cointegrating eqn(s) at the 0.05 level

* denotes rejection of the hypothesis at the 0.05 level
**MacKinnon-Haug-Michelis (1999) p-values

H0: r=0
H1: 0 < r ≤ g
Reject H0, there are 2 cointegrating equations. We can go to ECM model.
Dependent Variable: D(USTB3M)
Method: Least Squares
Date: 12/17/23 Time: 23:11
Sample (adjusted): 8/03/2001 12/31/2009
Included observations: 2101 after adjustments

Variable Coefficient Std. Error t-Statistic Prob.

ECT1(-1) -0.044039 0.020000 -2.201903 0.0278

ECT2(-2) -0.017689 0.010430 -1.696009 0.0900

R-squared 0.023103 Mean dependent var -0.001614

Adjusted R-squared 0.022637 S.D. dependent var 0.064926
S.E. of regression 0.064187 Akaike info criterion -2.653070
Sum squared resid 8.647915 Schwarz criterion -2.647691
Log likelihood 2789.050 Hannan-Quinn criter. -2.651100
Durbin-Watson stat 1.687487

10
Assignment 2 and 3 Financial Econometrics (20%)

Dependent Variable: D(USTB6M)

Method: Least Squares
Date: 12/17/23 Time: 23:14
Sample (adjusted): 8/09/2001 12/31/2009
Included observations: 2097 after adjustments

Variable Coefficient Std. Error t-Statistic Prob.

D(USTB3M(-1)) 0.337370 0.032356 10.42667 0.0000

D(USTB3M(-2)) -0.202435 0.036596 -5.531625 0.0000
D(USTB3M(-3)) -0.173322 0.036926 -4.693819 0.0000
D(USTB3M(-4)) -0.061133 0.036465 -1.676474 0.0938
D(USTB3M(-5)) -0.204575 0.034746 -5.887790 0.0000
D(USTB3M(-6)) 0.000880 0.033485 0.026288 0.9790
D(USTB4W(-1)) -0.013097 0.016035 -0.816751 0.4142
D(USTB4W(-2)) 0.051734 0.017117 3.022395 0.0025
D(USTB4W(-3)) 0.028399 0.017411 1.631072 0.1030
D(USTB4W(-4)) 0.100621 0.016969 5.929820 0.0000
D(USTB4W(-5)) -0.018845 0.016551 -1.138562 0.2550
D(USTB4W(-6)) -0.003918 0.015836 -0.247426 0.8046
D(USTB6M(-1)) -0.261777 0.035863 -7.299279 0.0000
D(USTB6M(-2)) 0.065782 0.036992 1.778281 0.0755
D(USTB6M(-3)) 0.076167 0.036966 2.060466 0.0395
D(USTB6M(-4)) 0.114157 0.036888 3.094680 0.0020
D(USTB6M(-5)) 0.206599 0.036657 5.636074 0.0000
D(USTB6M(-6)) 0.020372 0.035099 0.580408 0.5617
C -0.001378 0.001020 -1.350845 0.1769
ECT1(-1) 0.046477 0.018902 2.458841 0.0140
ECT2(-1) -0.032334 0.010142 -3.188089 0.0015

R-squared 0.152808 Mean dependent var -0.001483

Adjusted R-squared 0.144646 S.D. dependent var 0.050347
S.E. of regression 0.046564 Akaike info criterion -3.286012
Sum squared resid 4.501207 Schwarz criterion -3.229448
Log likelihood 3466.383 Hannan-Quinn criter. -3.265293
F-statistic 18.72237 Durbin-Watson stat 2.003196
Prob(F-statistic) 0.000000

WALD Test.

11
Assignment 2 and 3 Financial Econometrics (20%)

If USTB3M Granger causes USTB6M.

c(1)=c(2)=c(3)=c(4)=c(5)=c(6)=0.
Wald Test:
Equation: Untitled

Test Statistic Value df Probability

F-statistic 36.57576 (6, 2076) 0.0000

Chi-square 219.4546 6 0.0000

Null Hypothesis: C(1)=C(2)=C(3)=C(4)=C(5)=C(6)=0

Null Hypothesis Summary:

Normalized Restriction (= 0) Value Std. Err.

C(1) 0.337370 0.032356

C(2) -0.202435 0.036596
C(3) -0.173322 0.036926
C(4) -0.061133 0.036465
C(5) -0.204575 0.034746
C(6) 0.000880 0.033485

Restrictions are linear in coefficients.

Wald Test:
Equation: Untitled

Test Statistic Value df Probability

F-statistic 36.57576 (6, 2076) 0.0000

Chi-square 219.4546 6 0.0000

Null Hypothesis: C(1)=C(2)=C(3)=C(4)=C(5)=C(6)=0

Null Hypothesis Summary:

Normalized Restriction (= 0) Value Std. Err.

C(1) 0.337370 0.032356

C(2) -0.202435 0.036596
C(3) -0.173322 0.036926
C(4) -0.061133 0.036465
C(5) -0.204575 0.034746
C(6) 0.000880 0.033485

Restrictions are linear in coefficients.

H0: USTB3M Granger cause USTB6M

H1: USTB3M Granger cause USTB6M
P-value less than 0.05, reject H0.

12
Assignment 2 and 3 Financial Econometrics (20%)

If USTB4W Granger cause USTB6M.

c(7)=c(8)=c(9)=c(10)=c(11)=c(12)=0.

Wald Test:
Equation: EQ02

Test Statistic Value df Probability

F-statistic 8.672013 (6, 2076) 0.0000

Chi-square 52.03208 6 0.0000

Null Hypothesis: C(7)=C(8)=C(9)=C(10)=C(11)=C(12)=0.

Null Hypothesis Summary:

Normalized Restriction (= 0) Value Std. Err.

C(7) -0.013097 0.016035

C(8) 0.051734 0.017117
C(9) 0.028399 0.017411
C(10) 0.100621 0.016969
C(11) -0.018845 0.016551
C(12) -0.003918 0.015836

Restrictions are linear in coefficients.

H0: USTB4W Granger Cause USTB6M.

H1: USTB4W, not Granger Cause USTB6M.
P-value less than 0.05, reject H0.

13
Assignment 2 and 3 Financial Econometrics (20%)

Variance Decomposition of USTB3M:

Period S.E. USTB3M USTB4W USTB6M
In the first period, the shock of USTB3M
1 0.059974 100.0000 0.000000 0.000000
2 0.091720 99.46794 0.478263 0.053801
comes from itself 100%. However, after ten
3 0.111451 97.81058 2.152248 0.037169 periods, the shock of USTB3M comes from
4 0.122610 97.12853 2.630415 0.241055 itself by 95%, from USTB4W by 1.5%, and
5 0.134472 96.37991 2.654959 0.965131 from USTB6M by 3%.
6 0.146634 95.54517 2.264474 2.190354
7 0.158321 95.24519 1.958098 2.796713
8 0.168316 95.20178 1.787293 3.010931
9 0.178057 95.27687 1.663642 3.059489
10 0.187466 95.29710 1.572986 3.129917

Variance Decomposition of USTB4W:

Period S.E. USTB3M USTB4W USTB6M In the first period, the shock of USTB4W
1 0.082658 40.16151 59.83849 0.000000
comes from USTB3M by 40%, and from
2 0.128965 52.96169 46.84683 0.191482 itself by 46%. After ten periods, the shock
3 0.159119 55.95519 43.51860 0.526217 comes from USTB3M by 67% and by
4 0.174399 56.87312 42.68278 0.444101
USTB4W by 32%.
5 0.187099 57.77993 41.79238 0.427686
6 0.198984 57.68567 41.81986 0.494477
7 0.210537 60.23364 39.30801 0.458351
8 0.220968 62.98057 36.60163 0.417804
9 0.230590 65.42062 34.19104 0.388340
10 0.239172 67.36125 32.27329 0.365458

Variance Decomposition of USTB6M:

Period S.E. USTB3M USTB4W USTB6M

1 0.046556 61.80189 0.572193 37.62592 In the first period, the shock of USTB6M
2 0.069243 73.25701 0.506259 26.23673 comes from USTB3M by 61%, and from
3 0.083583 73.54924 0.348236 26.10252
4 0.093608 71.49359 0.279443 28.22697
USTB6M by 37%. After ten periods, the
5 0.105139 69.77304 0.231987 29.99498 shock comes from USTB3M by 63% and
6 0.116686 66.94912 0.339331 32.71155 from USTB6M by 35%.
7 0.127392 65.49360 0.653736 33.85266
8 0.136734 64.21318 1.103675 34.68315
9 0.145398 63.66250 1.363727 34.97378
10 0.153723 63.20564 1.595185 35.19918

Cholesky One S.D. (d.f. adjusted)

Cholesky ordering: USTB3M USTB4W USTB6M

14
Assignment 2 and 3 Financial Econometrics (20%)

Question 2

Download the time series data of Maybank and Sunway REIT from Jan 1, 2020, to Nov 30,
2021. (download from Yahoo finance)
1. Compute the portfolio return and portfolio risk for 2020 and 2021. Is 2021 riskier
than 2020?
𝑅𝑝,𝑡 = ∑𝑛𝑖=1 𝑤𝑖 𝑅𝑖𝑡
2020 2021
Portfolio return -0.0152263 % 0.005349 %

Portfolio risk 1.6682289 % 0.812042 %

The portfolio risk in 2020 is 1.6682289 % which is higher than the portfolio risk in
2021 which is 0.812042 %. Therefore, we can conclude that 2020 is riskier than
2021.

2. Download KLCI data from Jan 1, 2020, to Nov 30, 2021

Conduct the test on the alpha and beta of the market model below.

R = α+βKLCI + ε

https://www.jstor.org/stable/2327592

Maybank Sunway REIT

Alpha, 𝛼 0.000298 -0.00031

Beta, 𝛽 0.933254 0.49791

The alpha for Maybank stock is higher than the alpha for Sunway REIT stock. This means that
Maybank stock is providing higher risk-adjusted return as compared to Sunway REIT stock.
A negative alpha for Sunway REIT means that the manager actually did worse than they should
have given the required return of the portfolio.
The beta for Maybank is higher than the beta for Sunway REIT which is 0.933254. The value is
near 1, which means that Maybank stock on average has a similar risk as compared to the market.
Sunway REIT has a beta of less than 1 which means that the asset is less volatile than the market
and less risky than the average market, and potentially to get lower return.

15
Assignment 2 and 3 Financial Econometrics (20%)

Question 3

Draw a combo chart to depict the growth of three variables. After that, please provide your
comment on the trend.
Islamic Asset Under Management for Malaysia Fund Management Industry
2018 2019 2020 2021
Total AUM (RM Bil) 743.58 823.19 905.46 968.76
Islamic AU 158.83 180.52 216.8 228.25
Percentage 21.36 21.93 23.94 23.56
Source: Securities Commission, Malaysia

Islamic Asset Under Management for Malaysia

Fund Management Industry
1200 24.5
24
1000
23.5
800 23
22.5
600
22
400 21.5
21
200
20.5
0 20
2018 2019 2020 2021

Total AUM (RM Bil) Islamic AUM Percentage

There is an increasing trend every year in the Islamic Asset Under Management for the Malaysia
Fund Management Industry.

16
Assignment 2 and 3 Financial Econometrics (20%)

Assignment 3 (10%)
State clearly SUBJECT CODE, your name, matric number, and email address on the Top
Right corner of the First page

1. A researcher, Rojen intends to use the CAPM model to estimate the risk and return
relationship of stock listed in the Bursa Malaysia.

The expected return of a given asset 𝑖 in the universe of all investment opportunities, according
to the Sharpe–Lintner CAPM model, should be equal to Eq. (1).

𝐸(𝑅𝑖) = 𝑅𝑓 + [𝐸(𝑅𝑀) − 𝑅𝑓] 𝐵𝑖𝑀, (1)

Where 𝐸(𝑅𝑖)is the expected return of the individual asset 𝑖, which should be a risk-free rate 𝑅𝑓
plus the risk premium per unit of beta risk[𝐸(𝑅𝑀) − 𝑅𝑓] multiplied with 𝐵𝑖𝑀. The beta coefficient
of the individual asset return, 𝐵𝐼𝑀 is the covariance risk of asset 𝑖 relative to the average
covariance of all risk assets.

(a) Rewrite the theoretical model above into an empirical model. Identify the parts of the
equation.

𝑅𝑖 − 𝑟𝑓 = 𝛼 + 𝛽(𝑅𝑚 − 𝑟𝑓) + 𝜀

𝑅𝑖 = return to asset i.
𝑟𝑓 = risk-free rate.
𝑅𝑚 = market return.
𝛼 = intercept term
𝛽 = beta of the asset i is the covariance of its return with the market return divided by the
variance of the market return.
𝜀 = random return for stock i.
(𝑅𝑚 − 𝑟𝑓) = market risk premium.

17
Assignment 2 and 3 Financial Econometrics (20%)

(b) Rojen has the data from Jan 5, 2021, to April 20, 2023. Suppose he intends to see
whether the Beta has changed after the General Election on November 19, 2022. Restate the
model using a dummy variable and identify the sub-sample period clearly.

Let Dt be the dummy variable that takes the value 0 for the period before the election (January 5,
2021, to November 19, 2022) and 1 for the period after the election (November 20, 2022, to April
20, 2023).
The sub-sample period:
Before the general election: January 5, 2021, to November 19, 2022. (Dummy variable Dt = 0)
After the general election: November 20, 2022, to April 20, 2023. (Dummy variable Dt = 1)

𝐸(𝑅𝑖) = 𝑅𝑓 + [𝐸(𝑅𝑀) − 𝑅 𝑓] 𝐵𝑖𝑀

𝐸(𝑅𝑖 ) = 𝑅𝑓 + 𝛽𝑖 (𝐸(𝑅𝑚 ) − 𝑅𝑓 ) + 𝛿𝐷𝑖 + 𝜀𝑖

𝑟𝑖 = 𝛼𝑖 + 𝛽𝑖 𝑟𝑚 + 𝛾𝑖 𝐷𝑡 𝑟𝑚 + 𝜀

18
Assignment 2 and 3 Financial Econometrics (20%)

2. Based on the idea of a market model proposed by Stapleton and Subrahmanyam (1983) in
the Journal of Finance, Vol. 38, No.5, as shown below, where 𝑅𝑖 is the return of the individual
asset 𝑖, and 𝑅𝑀 is the return of the market benchmark.

(a) Based on the estimated output below, test whether the stock is aggressive or defensive.

Hypothesis:

H0: β = 1, the stock is aggressive.

H1: β < 1, the stock is defensive.

Decision:

From the regression output, the coefficient of the β = 0.129061 which is less than 1.
Therefore, we must reject the null hypothesis that the stock is aggressive. We can
conclude that the stock is defensive, or the stock doesn’t move much because the
coefficient of β is less than 1.

19
Assignment 2 and 3 Financial Econometrics (20%)

(b) Based on the estimated output below, conduct a test to ascertain whether the return stock
has a serial correlation. Suggest a method to mitigate the problem (if there is).

H0: There is no autocorrelation.

H1: There is autocorrelation.

The value of Durbin-Watson statistic is 2.009385 which it is quite close to the expected value of
2. This suggests that there is not strong evidence of serial correlation in the residuals. Therefore,
we do not reject the null hypothesis of no autocorrelation.

20
Assignment 2 and 3 Financial Econometrics (20%)

(c) Based on the estimated output below, conduct a test to ascertain whether the return stock
has heteroscedasticity. Suggest a method to mitigate the problem (if there is one).

H0: Homoskedastic.
H1: Heteroskedasticity.
P-value = 0.2781 > a = 0.05. Do not reject H0 at 5% significance level, there is no
heteroskedasticity on the return stock.

21
Assignment 2 and 3 Financial Econometrics (20%)

(d) Based on the estimated output below, conduct a test to ascertain whether the beta parameter
has changed after the General Election held on 19 November 2022.

H0: No breaks at specified breakpoints (coefficients are same in both groups).

H1: There’s break at specifies breakpoints.
P-value = 0.6768 > a = 0.05, do not reject H0. There is no breakpoint on 19 November 2022.

22