Você está na página 1de 8

TESTING THE STATIONARITY

ADF Level

Philips-Person (Level)

RKse

-9.655912
-9.655912

RUS

-3.870720

-9.268799

-7.750283

RAus

-7.948884

Critcal VAlues
1%

-3.486551

-3.486551

-2.886074

5%

-2.886074
-2.579931

10%

-2.579931

DATA IS STATIONARY AND ALSO CONFIRM THROUGH FLIP PREN TEST


AS SHOWN BY THE ABOVE TABLE.
CALCULATION OF LEG
VAR Lag Order Selection Criteria
Endogenous variables: RKSE RUS RAUS
Exogenous variables: C
Date: 08/14/12 Time: 13:05
Sample: 1 119
Included observations: 113
Lag

LogL

LR

FPE

AIC

SC

HQ

0
1
2
3
4
5
6

509.7821
522.0118
536.9514
542.9101
547.8164
555.1360
562.0311

NA
23.59344
28.02831*
10.86282
8.683652
12.56651
11.47142

2.55e-08
2.41e-08
2.17e-08*
2.29e-08
2.47e-08
2.55e-08
2.66e-08

-8.969595
-9.026757
-9.131883*
-9.078055
-9.005600
-8.975859
-8.938604

-8.897187*
-8.737123
-8.625023
-8.353970
-8.064289
-7.817323
-7.562842

-8.940213*
-8.909226
-8.926205
-8.784229
-8.623625
-8.505737
-8.380334

* indicates lag order selected by the criterion


LR: sequential modified LR test statistic (each test at 5% level)
FPE: Final prediction error
AIC: Akaike information criterion
SC: Schwarz information criterion
HQ: Hannan-Quinn information criterion

MODEL SPICISFICATION

M2
Trace
Critical
Statistic
Value
0 107.4995 35.19275
1 48.18301 20.26184
2 19.05119 9.164546

M3
Trace
Critical
Statistic
Value
107.4513 29.79707
48.13826 15.49471
19.03726 3.841466

M4
Trace
Critical
Statistic
Value
110.386 42.91525
49.76324 25.87211
20.25298 12.51798

M5
Trace
Critical
Statistic
Value
110.2756 35.0109
49.71136 18.39771
20.20131 3.841466

In this case I suggest model 3 to use although the above result show that from model 2 to model 5 value
of t-statistic is greater then critical values because empirical study show most research paper use model
three.

Pairwise Granger Causality Tests


Date: 08/14/12 Time: 13:32
Sample: 1 119
Lags: 2
Null Hypothesis:

Obs

F-Statistic

Prob.

RUS does not Granger Cause RKSE


RKSE does not Granger Cause RUS

117

7.22532
2.91746

0.0011
0.0582

RAUS does not Granger Cause RKSE


RKSE does not Granger Cause RAUS

117

8.27552
4.57851

0.0004
0.0123

RAUS does not Granger Cause RUS


RUS does not Granger Cause RAUS

117

2.31443
0.26728

0.1035
0.7659

table indicate that US and KSE has unisdirectional lead lag relationship.
AUSTIRIA and KSE have bidirectional lead lag relationship.

IMPLUS RESOPONSE

Response to Cholesky One S.D. Innovations


Response of RKSE to RUS
.025
.020
.015
.010
.005
.000
-.005
1

10

10

Response of RKSE to RAUS


.025
.020
.015
.010
.005
.000
-.005
1

The above figures showas that there is more respond in kse when us market changes. But Austria
market less influence kse.

DE COMPOSIATION TEST

Period

S.E.

RKSE

RUS

RAUS

1
2
3
4
5
6
7
8
9
10

0.084008
0.085942
0.091011
0.091326
0.091831
0.091881
0.091980
0.091990
0.092001
0.092002

100.0000
95.82357
85.70561
85.25372
85.33863
85.28315
85.11466
85.09905
85.10017
85.09936

0.000000
4.171239
8.991940
8.938444
8.910138
8.969682
9.059049
9.058145
9.058137
9.059080

0.000000
0.005194
5.302445
5.807841
5.751234
5.747167
5.826295
5.842805
5.841697
5.841562

Choles
ky
Orderin
g:
RKSE
RUS
RAUS

the above table shows that more change in kse is due to internal factors of kse as shown by 2nd period to
10 period(95% to 85%) but us and aus does not more influence kse.

Autoregressive Distributed Lag Estimates


ARDL(0,0,2) selected based on Schwarz Bayesian Criterion
****************************************************************
Dependent variable is X1
115 observations used for estimation from
3 to 117
Regressor
Coefficient
Standard Error
T-Ratio[Prob]
X2
-.20368
.29872
-.68183[.497]
X3
.37490
.16462
2.2774[.025]
X3(-1)
.12026
.12655
.95027[.344]
X3(-2)
.32055
.12235
2.6200[.010]
***************************************************************
R-Squared
.14685
R-Bar-Squared
.12379
S.E. of Regression
.084139
F-stat
F(3,111)
6.3688[.001]
Mean of Dependent Vari-.018085 S.D. of Dependent Variable .089887
Residual Sum of Squares.78582
Equation Log-likelihood 123.5149
Akaike Info. Criterion 119.
Schwarz Bayesian Criterion 114.0250
DW-statistic
1.8709

THE ABOVE table indicates that statistical value of x2(us market) is more than .05 so it has insignificant
relationship with kse(x1). And x3(AUSTRIA) has significant relationship with kse(x1).

Estimated Long Run Coefficients using the ARDL Approach


ARDL(0,0,2) selected based on Schwarz Bayesian Criterion
*****************************************************************************
**
Dependent variable is X1
115 observations used for estimation from
3 to 117
*****************************************************************************
**
Repressor
Coefficient
Standard Error
T-Ratio[Prob]
X2
-.20368
.29872
-.68183[.497]
X3
.81572
.18318
4.4532[.000]
*****************************************************************************
**

THE ABOVE table indicates that statistical value of x2(us market) is more than .05 so it has long run
insignificant relationship with kse(x1). And x3(AUSTRIA) has long run significant relationship with
kse(x1).

ERROR CORRECTION MODEL


Error Correction Representation for the Selected ARDL Model
ARDL(0,0,2) selected based on Schwarz Bayesian Criterion
*****************************************************************************
**
Dependent variable is dX1
115 observations used for estimation from
3 to 117
*****************************************************************************
**
Repressor
Coefficient
Standard Error
T-Ratio[Prob]
dX2
-.20368
.29872
-.68183[.497]
dX3
.37490
.16462
2.2774[.025]
dX31
-.32055
.12235
-2.6200[.010]
ecm (-1)
-1.0000
0.00
0.0001

THE CO OFFFICENT OF ECM TERM HAS CORRECT SIGHN ANC SIGNIFICANT IT CONFIRMS PRESENCES OF
RELATIONSHIP BETWEEN VARIABLES. THE COEFFENT ECM SUGGEST THATADJUSTEMENTS PROESS IS.1
OF THE PRIVIOUS YEAR DISEQULIBRIM KSE(X1) FROM ITS DIS EQULIBRIM IN THE CURRRENT YEAR

Plot of Cumulative Sum of Recursive


Residuals
30
25
20
15
10
5
0
-5
-10
-15
-20
-25
-30
3

13

23

33

43

53

63

73

83

93

103

The straight lines represent critical bounds at 5% significance level

113

117

Plot of Cumulative Sum of Squares of


Recursive Residuals
1.5
1.0
0.5
0.0
-0.5
3

13

23

33

43

53

63

73

83

93

103

113

117

The straight lines represent critical bounds at 5% significance level

THE ABOVE GRAOPH SHOWS BOTH STATISTAIS CUSAUM AND CUSAUM SQURE ARE WITH IN THE
CRITICAL BOUNDS.05% INDICATE THE MODLE IS STRIUCLRLY STABLE

Você também pode gostar