Mean Reversion in Swedish Macroeconomic Time Series - Evidence Using a New Panel Data Approach
2004 (English)In: European Review of Economics and Finance, Vol. 3, no 4, 23-36 p.Article in journal (Refereed) Published
The presence or absence of unit roots in a time series can be used to test the validity of a number of economic hypotheses and the persistence of economic time series therefore receives a fair amount of attention. This paper tests for the presence of unit roots in four time series of major interest to the Swedish macro economy. The time series properties of the real exchange rate, the nominal interest rate, inflation and unemployment are investigated using both traditional univariate unit root tests and panel unit root tests in a new panel setting. It is well known that there is power to be gained when testing for unit roots by using a panel setting. By applying two different panel unit root tests - the frequently used Im, Pesaran and Shin (2003) and the less used, but potentially highly informative, Johansen (1988) likelihood ratio test - the drawbacks of panel unit root tests, such as formulation of null and alternative hypothesis and the common assumption of cross-sectional independence, are also addressed in this study. Applying the tests to monthly data from 1972 to 2003, the panel unit root tests provide strong evidence that all four time series are stationary-processes.
Place, publisher, year, edition, pages
2004. Vol. 3, no 4, 23-36 p.
Panel data, Unit root tests
IdentifiersURN: urn:nbn:se:uu:diva-68306OAI: oai:DiVA.org:uu-68306DiVA: diva2:96217