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



ci rel


20 г-                            Polynomial cointegration relation 1


1980                1985                1990                1995


1980                1985                1990                1995


-10


3∙pdf                     1980              1985               1990              1995

Figure 2: The graphs of the three polynomial cointegration relations. Upper panel
describes the IKE relation, the middel panel the inflation expectations relation, and
the lower panel the German inflation rate relation.

β'2xt + δ'2xt = {- s2) + 0.60∆p2 - O.51∆pι - O.17∆sι2 + 0.00sι2 + 0.00⅛.1} (23)

The third relation, essentially a relation for German inflation rate, is similar to the
relation found in Juselius and MacDonald (2006) and describes the latter as (almost)
homogeneously related to US inflation rate, German short-term interest rate, and the
change in the Dmk/$ rate:

β3xt + J3xt = {1.31∆p1 - 0.31∆p2 - 0.74s1 - 0.07∆s12 + 0.00pp - 0.00t91.1} (24)

All three relations contain a tiny, but significant, trend effect which is more difficult
to interpret. The most likely explanation is the usual one that the linear trend effect in
the relations is a proxy for some information not included in the analysis. For example,
the small trend effect in (22) might account for some perceived productivity differential
between the two economies. In (23) the re-unification trend might be a proxy for a
change in the market’s re-assessment of the riskiness of the nominal Dmk/$ rate. In
(24) the trend together with the
pp may imply that German inflation rate, in addition
to following the US inflation rate, the short-term interest rate, and the change in the
Dmk/$ rate, has exhibited a long-run adjustment to trend-adjusted relative prices.
Figure 2 shows that the three polynomially cointegrating relations are very stationary.
!

The β±11xt relations are CI(2,1) cointegrating relations which only become sta-

'

tionary by differencing. Thus, βx1xt could be interpretable as partially specified

20



More intriguing information

1. Party Groups and Policy Positions in the European Parliament
2. Regional differentiation in the Russian federation: A cluster-based typification
3. Evaluating the Impact of Health Programmes
4. The migration of unskilled youth: Is there any wage gain?
5. THE INTERNATIONAL OUTLOOK FOR U.S. TOBACCO
6. The name is absent
7. The demand for urban transport: An application of discrete choice model for Cadiz
8. The name is absent
9. INTERPERSONAL RELATIONS AND GROUP PROCESSES
10. The name is absent
11. The name is absent
12. Modelling the health related benefits of environmental policies - a CGE analysis for the eu countries with gem-e3
13. The name is absent
14. The Impact of Minimum Wages on Wage Inequality and Employment in the Formal and Informal Sector in Costa Rica
15. The name is absent
16. A Bayesian approach to analyze regional elasticities
17. The name is absent
18. Backpropagation Artificial Neural Network To Detect Hyperthermic Seizures In Rats
19. Estimating the Technology of Cognitive and Noncognitive Skill Formation
20. WP 48 - Population ageing in the Netherlands: Demographic and financial arguments for a balanced approach