Tải bản đầy đủ
1 Appendix-1: Testing for ARCH Effect

1 Appendix-1: Testing for ARCH Effect

Tải bản đầy đủ

Master of Business Administration

Lam Van Bao Dan

6.1.2 Crisis Period (From January, 2008 to December, 2009)
Heteroskedasticity Test: ARCH
F-statistic
Obs*R-squared

23.52759
22.54512

Prob. F(1,492)
Prob. Chi-Square(1)

0.0000
0.0000

Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 03/24/12 Time: 17:32
Sample (adjusted): 3 496
Included observations: 494 after adjustments
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
RESID^2(-1)

0.000357
0.213541

3.32E-05
0.044024

10.76368
4.850525

0.0000
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

K17-EMBA

0.045638
0.043698
0.000589
0.000171
2973.643
23.52759
0.000002

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

Page 66

0.000454
0.000603
-12.03094
-12.01393
-12.02426
2.016806

2012

Master of Business Administration

Lam Van Bao Dan

6.1.3 Recovering Period (From January, 2010 to December, 2011)
Heteroskedasticity Test: ARCH
F-statistic
Obs*R-squared

22.00790
21.15456

Prob. F(1,494)
Prob. Chi-Square(1)

0.0000
0.0000

Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 03/24/12 Time: 17:39
Sample (adjusted): 3 498
Included observations: 496 after adjustments
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
RESID^2(-1)

0.000133
0.207738

1.40E-05
0.044282

9.506591
4.691258

0.0000
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

K17-EMBA

0.042650
0.040712
0.000265
3.46E-05
3382.450
22.00790
0.000004

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

Page 67

0.000167
0.000270
-13.63085
-13.61389
-13.62419
2.049218

2012

Master of Business Administration

Lam Van Bao Dan

6.1.4 Whole Period (From March, 2002 to December, 2011)
Heteroskedasticity Test: ARCH
F-statistic
Obs*R-squared

362.1974
315.6900

Prob. F(1,2443)
Prob. Chi-Square(1)

0.0000
0.0000

Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 03/24/12 Time: 17:42
Sample (adjusted): 3 2447
Included observations: 2445 after adjustments
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
RESID^2(-1)

0.000149
0.359316

9.71E-06
0.018880

15.31713
19.03149

0.0000
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)

K17-EMBA

0.129117
0.128760
0.000428
0.000448
15494.02
362.1974
0.000000

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat

Page 68

0.000232
0.000459
-12.67241
-12.66766
-12.67068
2.077143

2012

Master of Business Administration

6.2

Lam Van Bao Dan

Appendix-2: GARCH Models Analysis

6.2.1 Before Crisis Period (From March, 2002 to December, 2007)
ARCH (1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:05
Sample (adjusted): 2 1453
Included observations: 1452 after adjustments
Convergence achieved after 12 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(2) + C(3)*RESID(-1)^2
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

-9.53E-05

0.000210

-0.453578

0.6501

36.73212
11.69130

0.0000
0.0000

Variance Equation
C
RESID(-1)^2
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

7.31E-05
0.739469
-0.005117
-0.005117
0.013786
0.275762
4405.753
1.494410

1.99E-06
0.063249

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 69

-0.001079
0.013751
-6.064398
-6.053488
-6.060327

2012

Master of Business Administration

Lam Van Bao Dan

GARCH (1,1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:07
Sample (adjusted): 2 1453
Included observations: 1452 after adjustments
Convergence achieved after 12 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(2) + C(3)*RESID(-1)^2 + C(4)*GARCH(-1)
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

0.000256

0.000135

1.904324

0.0569

6.705529
12.12539
40.58187

0.0000
0.0000
0.0000

Variance Equation
C
RESID(-1)^2
GARCH(-1)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

2.77E-06
0.385811
0.674700
-0.009429
-0.009429
0.013815
0.276945
4642.188
1.488026

4.12E-07
0.031818
0.016626

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 70

-0.001079
0.013751
-6.388688
-6.374141
-6.383260

2012

Master of Business Administration

Lam Van Bao Dan

TGARCH (1,1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:08
Sample (adjusted): 2 1453
Included observations: 1452 after adjustments
Convergence achieved after 16 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(2) + C(3)*RESID(-1)^2 + C(4)*RESID(-1)^2*(RESID(-1)<0) +
C(5)*GARCH(-1)
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

0.000351

0.000158

2.217178

0.0266

6.548112
9.383390
-2.315439
41.32354

0.0000
0.0000
0.0206
0.0000

Variance Equation
C
RESID(-1)^2
RESID(-1)^2*(RESID(-1)<0)
GARCH(-1)

2.63E-06
0.466888
-0.134241
0.674133

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

-0.010809
-0.010809
0.013825
0.277324
4644.298
1.485994

K17-EMBA

4.02E-07
0.049757
0.057976
0.016314

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 71

-0.001079
0.013751
-6.390218
-6.372033
-6.383432

2012

Master of Business Administration

Lam Van Bao Dan

EGARCH (1,1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:09
Sample (adjusted): 2 1453
Included observations: 1452 after adjustments
Convergence achieved after 16 iterations
Presample variance: backcast (parameter = 0.7)
LOG(GARCH) = C(2) + C(3)*ABS(RESID(-1)/@SQRT(GARCH(-1))) + C(4)
*RESID(-1)/@SQRT(GARCH(-1)) + C(5)*LOG(GARCH(-1))
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

0.000283

0.000160

1.775311

0.0758

-14.51004
17.58108
1.396136
143.5665

0.0000
0.0000
0.1627
0.0000

Variance Equation
C(2)
C(3)
C(4)
C(5)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

-0.986306
0.596224
0.033311
0.941036
-0.009818
-0.009818
0.013818
0.277052
4647.058
1.487453

0.067974
0.033913
0.023860
0.006555

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 72

-0.001079
0.013751
-6.394020
-6.375836
-6.387235

2012

Master of Business Administration

Lam Van Bao Dan

GARCH-M (1,1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:04
Sample (adjusted): 2 1453
Included observations: 1452 after adjustments
Convergence achieved after 27 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(3) + C(4)*RESID(-1)^2 + C(5)*GARCH(-1)
Variable

Coefficient

Std. Error

z-Statistic

Prob.

GARCH
C

-4.324580
0.000451

1.666026
0.000150

-2.595747
2.997571

0.0094
0.0027

6.660993
12.19970
42.40696

0.0000
0.0000
0.0000

Variance Equation
C
RESID(-1)^2
GARCH(-1)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

2.67E-06
0.375240
0.682140
0.001476
0.000787
0.013745
0.273953
4647.294
1.520155

4.01E-07
0.030758
0.016086

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 73

-0.001079
0.013751
-6.394344
-6.376160
-6.387559

2012

Master of Business Administration

Lam Van Bao Dan

6.2.2 Crisis Period (From January, 2008 to December, 2009)
ARCH (1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:10
Sample (adjusted): 2 496
Included observations: 495 after adjustments
Convergence achieved after 12 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(2) + C(3)*RESID(-1)^2
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

0.000694

0.001012

0.685814

0.4928

9.429063
2.312012

0.0000
0.0208

Variance Equation
C
RESID(-1)^2
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

0.000395
0.232104
-0.000607
-0.000607
0.022803
0.256879
1179.115
1.293398

4.19E-05
0.100390

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 74

0.001255
0.022797
-4.751978
-4.726496
-4.741975

2012

Master of Business Administration

Lam Van Bao Dan

GARCH (1,1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:11
Sample (adjusted): 2 496
Included observations: 495 after adjustments
Convergence achieved after 33 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(2) + C(3)*RESID(-1)^2 + C(4)*GARCH(-1)
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

0.000994

0.000875

1.136424

0.2558

1.584444
2.526841
12.89560

0.1131
0.0115
0.0000

Variance Equation
C
RESID(-1)^2
GARCH(-1)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

1.76E-05
0.145088
0.824225
-0.000131
-0.000131
0.022798
0.256757
1194.484
1.294013

1.11E-05
0.057419
0.063915

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 75

0.001255
0.022797
-4.810036
-4.776060
-4.796698

2012

Master of Business Administration

Lam Van Bao Dan

TGARCH (1,1)
Dependent Variable: RETURN
Method: ML - ARCH (Marquardt) - Normal distribution
Date: 04/01/12 Time: 11:12
Sample (adjusted): 2 496
Included observations: 495 after adjustments
Convergence achieved after 94 iterations
Presample variance: backcast (parameter = 0.7)
GARCH = C(2) + C(3)*RESID(-1)^2 + C(4)*RESID(-1)^2*(RESID(-1)<0) +
C(5)*GARCH(-1)
Variable

Coefficient

Std. Error

z-Statistic

Prob.

C

0.000905

0.000905

1.000299

0.3172

1.356175
2.262782
1.188027
13.78114

0.1750
0.0236
0.2348
0.0000

Variance Equation
C
RESID(-1)^2
RESID(-1)^2*(RESID(-1)<0)
GARCH(-1)
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

K17-EMBA

1.53E-05
0.116660
0.083618
0.818565
-0.000237
-0.000237
0.022799
0.256784
1195.790
1.293877

1.13E-05
0.051556
0.070384
0.059397

Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.

Page 76

0.001255
0.022797
-4.811275
-4.768804
-4.794602

2012