Powered by OpenAIRE graph
Found an issue? Give us feedback
addClaim

Stochastic unit-root models in economics

Authors: Holl, Jürgen;

Stochastic unit-root models in economics

Abstract

Die vorliegende Dissertation besteht aus drei Essays, die von sogenannten stochastischen Einheitswurzel-(STUR-)Modellen in der Volkswirtschaftslehre handeln. Ein Test der Nullhypothese einer Einheitswurzel in einem autoregressiven Modell erster Ordnung (AR(1)) gegen Alternativen, die in einem STUR-Modell eingebettet sind, wird vorgestellt. STUR-Modelle ermöglichen den Wechsel zwischen stationären und explosiven Regimen. Der Pseudo-Likelihood-Quotienten-Test basiert auf der Äquivalenz der ersten und zweiten bedingten Momente von STUR-Modellen und AR(1)-Modellen mit Störtermen eines autoregressiven bedingt heteroskedastischen Modells erster Ordnung. Die asymptotische Verteilung wird hergeleitet und kritische Werte werden simuliert. Monte-Carlo-Experimente zeigen, dass der Test tatsächlich Teststärke gegen stationäre und STUR-Alternativen besitzt. Der Test wird auf die Arbeitslosenquoten von zehn Staaten angewandt. Unter Verwendung Bayesianischer Methoden werden STUR-Modelle für diese Arbeitslosenquoten gefittet, um die Entscheidungen von Einheitswurzel-Tests, die Teststärke gegen STUR-Prozesse besitzen, zu evaluieren. Das Bayesianische Verfahren ermöglicht die Einschätzung der Abweichung der STUR-Modelle von gewöhnlichen Zeitreihenmodellen mit konstanten Koeffizienten. Die geschätzten Werte für die Parameter sind hoch signifikant für alle Staaten. Da das Bayesianische Verfahren sehr zeitaufwendig ist, wird ein alternatives Verfahren in Simulationen getestet. Die Anwendung der STUR-Modelle auf Arbeitslosenquoten wird aus einer ökonomischen Perspektive motiviert. Prognosen der Arbeitslosenquoten, die sich aus STUR-Modellen errechnen, werden evaluiert. Resultate der Einheitswurzel-Tests angewandt auf reale und simulierte Daten werden den Resultaten von 1-Schritt-, 3-Schritt und 12-Schritt-Prognosen gegenübergestellt; STUR wird dabei mit drei gewöhnlichen Zeitreihenmodellen verglichen. Darüberhinaus werden kombinierte Prognosen berechnet. Die Frage, ob STUR überhaupt ein gutes Prognosemodell ist, wird erörtert. Test und Prognose führen nicht für jeden Staat zum selben Ergebnis, jedoch für bestimmte Staaten ist STUR ein relevantes Modell und sollte als echte Alternative betrachtet werden.

The dissertation at hand consists of three essays focussing on so-called stochastic unit-root (STUR) models in economics. A test of the null hypothesis of a unit root in an autoregressive model of order one (AR(1)) against alternatives nested in a STUR model is introduced. STUR models allow for changes between stationary and explosive regimes. The pseudo-likelihood ratio test is based on the equivalence in first and second conditional moment of STUR models and AR(1) models with errors from an autoregressive conditional heteroskedastic model of order one. Asymptotics are derived and critical values simulated. Monte Carlo experiments show that the test really has power against stationary and STUR alternatives. An application to unemployment rates of ten countries is provided. To evaluate the decisions of unit-root tests having power against STUR processes, STUR models are fitted to the unemployment rates by using Bayesian techniques. The Bayesian procedure in use allows to assess the deviation of STUR models from standard fixed-coefficient time series models. Parameter estimates show up strongly significant for all countries. As the Bayesian method is a very time-consuming one, an alternative procedure is studied in simulations. The application of STUR models to unemployment rates is motivated from an economic point of view. Forecasts of the unemployment rates generated from STUR models are evaluated. Unit-root test results on original and simulated data are confronted with 1-step, 3-step and 12-step forecast results where STUR is compared to three standard time series models. Additionally, combined forecasts are calculated. The question whether STUR is a good forecast model at all is addressed. Testing and forecasting do not coincide for every country. However, the discussion suggests that STUR is relevant for certain countries and should be considered a real alternative.

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Related to Research communities
Upload OA version
Are you the author of this publication? Upload your Open Access version to Zenodo!
It’s fast and easy, just two clicks!