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. AGRIBUSINESS EXECUTIVE EDUCATION AND KNOWLEDGE EXCHANGE: NEW MECHANISMS OF KNOWLEDGE MANAGEMENT INVOLVING THE UNIVERSITY, PRIVATE FIRM STAKEHOLDERS AND PUBLIC SECTOR
2. Reversal of Fortune: Macroeconomic Policy, International Finance, and Banking in Japan
3. Models of Cognition: Neurological possibility does not indicate neurological plausibility.
4. Target Acquisition in Multiscale Electronic Worlds
5. The name is absent
6. Evolution of cognitive function via redeployment of brain areas
7. Whatever happened to competition in space agency procurement? The case of NASA
8. Parent child interaction in Nigerian families: conversation analysis, context and culture
9. Demand Potential for Goat Meat in Southern States: Empirical Evidence from a Multi-State Goat Meat Consumer Survey
10. Innovation in commercialization of pelagic fish: the example of "Srdela Snack" Franchise