Journal Article
No. 2007-3 | May 21, 2007
Long Run Macroeconomic Relations in the Global Economy


This paper presents tests of long run macroeconomic relations involving interest rates, equity, prices and exchange rates suggested by arbitrage in financial and goods markets. It uses the global vector autoregressive (GVAR) model to test for long run restrictions in each country/region conditioning on the rest of the world. Bootstrapping is used to compute both the empirical distribution of the im¬pulse responses and the log-likelihood ratio statistic for over-identifying restrictions. The paper also examines the speed with which adjustments to the long run relations take place via the persistence profiles. It finds strong evidence in favour of a long run version of uncovered interest parity and to a lesser extent the Fisher equation across a number of countries, but the test results for the purchasing power parity relation are much weaker. Also the transmission of shocks and subsequent ad¬justments in financial markets are much faster than those in goods markets.

JEL Classification:

C32, E17, F47, R11



Cite As

Stephane Dees, Sean Holly, Hashem Pesaran, and L. Vanessa Smith (2007). Long Run Macroeconomic Relations in the Global Economy. Economics: The Open-Access, Open-Assessment E-Journal, 1 (2007-3): 1—58.

Comments and Questions

Anonymous - When 'Global' is in fact local
September 05, 2007 - 11:31

A complicated paper for a simple issue. Global economy seems to be Europe and Anglo-Saxon world (OK, China is added, too). Results are not clear cut and they seem to have been 'discovered' in the previous literature.

Anonymous - Reply by Dees, Holly, Pesaran and Smith to the comment "When 'Global' is in fact local"
November 07, 2007 - 13:54

Many thanks for the comment. But there are two points that require clarification. First, our global model (the GVAR) comprises 26 countries that account for more than 90% of world output. Contrary to what is stated it also covers countries in Latin America (5), South East Asia (5), Korea, India, ...[more]

... Saudi Arabia, Turkey and South Africa. Second, the type of questions we are asking from the model are far from simple!

Just to emphasize the contribution of the paper lies in:

(i) Showing how to test the validity of over-identifying restrictions implied by specific long run structural relations suggested by economic theory, within a global model. To obtain the critical values of the likelihood ratio statistics for testing the over-identifying restrictions requires bootstrapping from the solution of the global model as described in the paper, where the GVAR model is solved in terms of the 134 domestic variables in the world economy;
(ii) Showing how the GVAR model requires to be modified in order for long-run relations such as the Purchasing Power Parity to be imposed on the model;
(iii) Examining the dynamic properties of a more theoretically transparent global model and comparing them to those of an otherwise unrestricted global model as presented in DdPS (2007).

Over-identifying restrictions are tested for 11 countries namely Europe, the Anglo-Saxon world and China, while the remaining 15 individual country models are estimated subject to just-identifying restrictions. The number or choice of countries subject to over-identifying restrictions is by no means constrained. It is up to the modeler and depends on the purpose of the exercise. The individual country models are then estimated with the long run relations imposed - where applicable - and are linked in a consistent manner to provide a solution in terms of all 134 endogenous variables in the global economy. All results relating to impulse response analysis, persistence profiles and variance decomposition are therefore obtained for all 134 variables. Due to the large number of results only a subset of these are presented, the focus being on Europe, the Anglo-Saxon world and China, the countries with over-identifying restrictions. The results for the rest of the countries are available upon request.