Nonparametric cointegration analysis



Table A.6: Johansen’s test results for the number (r) of cointegrating vectors (intercept
present, but linear trend absent
)


p

r

test
stat.

crit.
10%

val.
5%

conclusions:

test         table(*)

type

10%

5%

_2_

__0__

___7._8__

__1__2.__1

_14__.0__

accept

accept

lambda-max     A.1

1

0.9

2.8

4.0

accept

accept

’’                     ’’

1

0.9

2.8

4.0

accept

accept

trace           ’’

0

8.7

13.3

15.2

accept

accept

’’                     ’’

0

7.8

12.8

14.6

accept

accept

lambda-max     A.2

1

0.9

6.7

8.1

accept

accept

’’                     ’’

1

0.9

6.7

8.1

accept

accept

trace           ’’

0

8.7

15.6

17.8

accept

accept

’’                     ’’

0

16.4

13.8

15.8

reject

reject

lambda-max     A.3

1

6.1

7.6

9.1

accept

accept

’’                     ’’

1

6.1

7.6

9.1

accept

accept

trace           ’’

0

22.5

18.0

20.2

reject

reject

’’                     ’’

1

8.63

2.71

3.84

reject

reject

interc. restr. χ2(1)

r=0

r=0

4

0

15.2

12.1

14.0

reject

reject

lambda-max     A.1

1

2.4

2.8

4.0

accept

accept

’’                     ’’

1

2.4

2.8

4.0

accept

accept

trace           ’’

0

17.6

13.3

15.2

reject

reject

’’                     ’’

0

15.2

12.8

14.6

reject

reject

lambda-max     A.2

1

2.4

6.7

8.1

accept

accept

’’                     ’’

1

2.4

6.7

8.1

accept

accept

trace           ’’

0

17.6

15.6

17.8

reject

accept

’’                     ’’

0

18.5

13.8

15.8

reject

reject

lambda-max     A.3

1

11.9

7.6

9.1

reject

reject

’’                     ’’

1

11.9

7.6

9.1

reject

reject

trace           ’’

0

30.5

18.0

20.2

reject

reject

’’                     ’’

1

3.36

2.71

3.84

reject

accept

interc. restr. χ2(1)

r=1

r=2

6

0

14.7

12.1

14.0

reject

reject

lambda-max     A.1

1

2.2

2.8

4.0

accept

accept

’’                     ’’

1

2.2

2.8

4.0

accept

accept

trace           ’’

0

16.9

13.3

15.2

reject

reject

’’                     ’’

0

14.7

12.8

14.6

reject

reject

lambda-max     A.2

1

2.2

6.7

8.1

accept

accept

’’                     ’’

1

2.2

6.7

8.1

accept

accept

trace           ’’

0

16.9

15.6

17.8

reject

accept

’’                     ’’

0

19.0

13.8

15.8

reject

reject

lambda-max     A.3

1

6.7

7.6

9.1

accept

accept

’’                     ’’

1

6.7

7.6

9.1

accept

accept

trace           ’’

0

25.7

18.0

20.2

reject

reject

’’                     ’’

1

4.31

2.71

3.84

reject

reject

interc. restr. χ2(1)

r=1

r=1


(*) Cf. Johansen and Juselius (1990). Table A.3 applies if cointegration restrictions have been imposed on
the intercept parameters, whereas tables A.1 and A.2 apply if no cointegration restrictions are imposed. Table
A.2 applies if these cointegration restrictions actually hold, and table A.1 applies if not. The
χ2(1) tests test the
null hypothesis that cointegration restrictions on the intercept parameters hold, given
r = 1, i.e., that the
cointegration relation contains an intercept rather than the error correction model itself.


62




More intriguing information

1. Ahorro y crecimiento: alguna evidencia para la economía argentina, 1970-2004
2. Insurance within the firm
3. The name is absent
4. The name is absent
5. Fortschritte bei der Exportorientierung von Dienstleistungsunternehmen
6. Second Order Filter Distribution Approximations for Financial Time Series with Extreme Outlier
7. Psychological Aspects of Market Crashes
8. Determinants of Household Health Expenditure: Case of Urban Orissa
9. Program Semantics and Classical Logic
10. NATURAL RESOURCE SUPPLY CONSTRAINTS AND REGIONAL ECONOMIC ANALYSIS: A COMPUTABLE GENERAL EQUILIBRIUM APPROACH