Testing Hypotheses in an I(2) Model with Applications to the Persistent Long Swings in the Dmk/$ Rate



Table 3: The trace test statistics for cointegration rank indices

p r

r

«2 = 7

«2 = 6

«2 = 5

«2 = 4

«2 = 3

«2 = 2

«2 = 1

«2 = 0

7

ɪ

1737.63

1402.12

1137.46

908.43

715.21

528.18

379.58

280.64

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

6

1

1115.86

885.59

659.57

470.53

319.31

190.76

160.82

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

[0.00]

5

2

646.64

455.20

271.26

141.15

94.27

93.63

[0.00]

[0.00]

[0.00]

[0.00]

[0.19]

[0.02]

4

3

265.40

134.94

69.81

39.63

40.67

[0.00]

[0.00]

[0.72]

[0.99]

[0.82]

3

4

64.60

32.60

15.58

22.18

[0.77]

[1.00]

[1.00]

[0.90]

2

5

13.04

2.03

8.06

[1.00]

[1.00]

[0.98]

1

6

0.91

0.17

[^A]

[1.00]

test the following four hypotheses:

1. H1 : R'1τ = 0, where R1 = [1,1, 0, 0, 0, 0, 0, 0, 0], i.e. we test whether we can
impose the
ppp restriction on all τ vectors. If accepted it would imply that the
nominal to real transformation
x't = [pppt, p1,t, p2,t, b1,t, b2,t, s1,t, s2,t] would
be econometrically valid (Kongsted, 2005) in the sense of transforming an
I(2)
vector to an I(1) without loss of information. The hypothesis is rejected based
on
y2(5) = 16.64 [0.01].

2. H2 : R12τ = 0, where R12 = [0, 0, 0, 0, 0, 0, 0,1,0], i.e. we test whether the broken
trend is long-run excludable from
τ. The hypothesis was borderline accepted
based
y2(5) = 9.99 [0.08] . Thus, there is only weak evidence that the direction of
the trend in relative prices and /or nominal exchange rates changed at the time
of the re-unification of Germany.

3. H3 : R3τ = 0, where R3 = [0, 0, 0, 0, 0, 0, 0, 0,1], i.e. we test whether the
3                         3

trend is long-run excludable from τ. The hypothesis was rejected based y2(5) =
13
.75 [0.02] . Thus, there is some evidence that the trend is econometrically needed
as a local approximation of the downward sloping trend in relative prices and in
the nominal exchange rate.

4.


H4 : R4τ = 0, where R14


0,0, 0, 0, 0,0, 0, 0,1

0,0, 0, 0, 0,0, 0,1, 0


, i.e. we test whether the


trends can be left out of the long-run relations. The hypothesis was rejected


based y2(10) = 24.50 [0.01] . Thus, the hypothesis that the downward sloping


trend strongly visible in Figure 1 is stochastic rather than deterministic is re-
jected. However, the hypotheses of no such trends were only borderline rejected,
which supports our prior assumption that a deterministic trend should only be
considered a local approximation.

17



More intriguing information

1. The name is absent
2. New Evidence on the Puzzles. Results from Agnostic Identification on Monetary Policy and Exchange Rates.
3. Education Research Gender, Education and Development - A Partially Annotated and Selective Bibliography
4. The name is absent
5. The InnoRegio-program: a new way to promote regional innovation networks - empirical results of the complementary research -
6. DISCRIMINATORY APPROACH TO AUDITORY STIMULI IN GUINEA FOWL (NUMIDA MELEAGRIS) AFTER HYPERSTRIATAL∕HIPPOCAMP- AL BRAIN DAMAGE
7. Policy Formulation, Implementation and Feedback in EU Merger Control
8. The name is absent
9. A parametric approach to the estimation of cointegration vectors in panel data
10. THE EFFECT OF MARKETING COOPERATIVES ON COST-REDUCING PROCESS INNOVATION ACTIVITY