What do we know about real exchange rate nonlinearities?

Research output: Contribution to journal › Article › Research › peer review

Authors

  • Robinson Kruse
  • Michael Frömmel
  • Lukas Menkhoff
  • Philipp Sibbertsen

Research Organisations

External Research Organisations

  • Aarhus University
  • Ghent University
View graph of relations

Details

Original languageEnglish
Pages (from-to)457-474
Number of pages18
JournalEmpirical economics
Volume43
Issue number2
Early online date4 Jan 2011
Publication statusPublished - Oct 2012

Abstract

Nonlinear modeling of adjustments to purchasing power parity has recently gained much attention. However, a huge body of the empirical literature applies ESTAR models and neglects the existence of other competing nonlinear models. Among these, the Markov Switching AR model has a strong substantiation in international finance. Our contribution to the literature is fivefold: First, ESTAR and MSAR models from a unit root perspective are compared. To this end, a new unit root test against MSAR is proposed as the second contribution. Thirdly, the case of misspecified alternatives in a Monte Carlo setup with real world parameter constellations is studied. The ESTAR unit root test is not indicative, while the MSAR unit test is robust. Fourthly, the case of correctly specified alternatives is considered and low power of the ESTAR but not for the MSAR unit root test is observed. Fifthly, an empirical application to real exchange rates suggests that they may indeed be explained by Markov Switching dynamics rather than ESTAR.

Keywords

    ESTAR, Markov Switching, PPP, Real exchange rates, Unit root test

ASJC Scopus subject areas

Cite this

What do we know about real exchange rate nonlinearities? / Kruse, Robinson; Frömmel, Michael; Menkhoff, Lukas et al.
In: Empirical economics, Vol. 43, No. 2, 10.2012, p. 457-474.

Research output: Contribution to journal › Article › Research › peer review

Kruse, R, Frömmel, M, Menkhoff, L & Sibbertsen, P 2012, 'What do we know about real exchange rate nonlinearities?', Empirical economics, vol. 43, no. 2, pp. 457-474. https://doi.org/10.1007/s00181-010-0431-2
Kruse, R., Frömmel, M., Menkhoff, L., & Sibbertsen, P. (2012). What do we know about real exchange rate nonlinearities? Empirical economics, 43(2), 457-474. https://doi.org/10.1007/s00181-010-0431-2
Kruse R, Frömmel M, Menkhoff L, Sibbertsen P. What do we know about real exchange rate nonlinearities? Empirical economics. 2012 Oct;43(2):457-474. Epub 2011 Jan 4. doi: 10.1007/s00181-010-0431-2
Kruse, Robinson ; Frömmel, Michael ; Menkhoff, Lukas et al. / What do we know about real exchange rate nonlinearities?. In: Empirical economics. 2012 ; Vol. 43, No. 2. pp. 457-474.
Download
@article{adafbaeb18a34765860c51db62987a97,
title = "What do we know about real exchange rate nonlinearities?",
abstract = "Nonlinear modeling of adjustments to purchasing power parity has recently gained much attention. However, a huge body of the empirical literature applies ESTAR models and neglects the existence of other competing nonlinear models. Among these, the Markov Switching AR model has a strong substantiation in international finance. Our contribution to the literature is fivefold: First, ESTAR and MSAR models from a unit root perspective are compared. To this end, a new unit root test against MSAR is proposed as the second contribution. Thirdly, the case of misspecified alternatives in a Monte Carlo setup with real world parameter constellations is studied. The ESTAR unit root test is not indicative, while the MSAR unit test is robust. Fourthly, the case of correctly specified alternatives is considered and low power of the ESTAR but not for the MSAR unit root test is observed. Fifthly, an empirical application to real exchange rates suggests that they may indeed be explained by Markov Switching dynamics rather than ESTAR.",
keywords = "ESTAR, Markov Switching, PPP, Real exchange rates, Unit root test",
author = "Robinson Kruse and Michael Fr{\"o}mmel and Lukas Menkhoff and Philipp Sibbertsen",
note = "Funding Information: Robinson Kruse gratefully acknowledges financial support from the CREATES funded by the Danish National Research Foundation. ",
year = "2012",
month = oct,
doi = "10.1007/s00181-010-0431-2",
language = "English",
volume = "43",
pages = "457--474",
journal = "Empirical economics",
issn = "0377-7332",
publisher = "Physica-Verlag",
number = "2",

}

Download

TY - JOUR

T1 - What do we know about real exchange rate nonlinearities?

AU - Kruse, Robinson

AU - Frömmel, Michael

AU - Menkhoff, Lukas

AU - Sibbertsen, Philipp

N1 - Funding Information: Robinson Kruse gratefully acknowledges financial support from the CREATES funded by the Danish National Research Foundation.

PY - 2012/10

Y1 - 2012/10

N2 - Nonlinear modeling of adjustments to purchasing power parity has recently gained much attention. However, a huge body of the empirical literature applies ESTAR models and neglects the existence of other competing nonlinear models. Among these, the Markov Switching AR model has a strong substantiation in international finance. Our contribution to the literature is fivefold: First, ESTAR and MSAR models from a unit root perspective are compared. To this end, a new unit root test against MSAR is proposed as the second contribution. Thirdly, the case of misspecified alternatives in a Monte Carlo setup with real world parameter constellations is studied. The ESTAR unit root test is not indicative, while the MSAR unit test is robust. Fourthly, the case of correctly specified alternatives is considered and low power of the ESTAR but not for the MSAR unit root test is observed. Fifthly, an empirical application to real exchange rates suggests that they may indeed be explained by Markov Switching dynamics rather than ESTAR.

AB - Nonlinear modeling of adjustments to purchasing power parity has recently gained much attention. However, a huge body of the empirical literature applies ESTAR models and neglects the existence of other competing nonlinear models. Among these, the Markov Switching AR model has a strong substantiation in international finance. Our contribution to the literature is fivefold: First, ESTAR and MSAR models from a unit root perspective are compared. To this end, a new unit root test against MSAR is proposed as the second contribution. Thirdly, the case of misspecified alternatives in a Monte Carlo setup with real world parameter constellations is studied. The ESTAR unit root test is not indicative, while the MSAR unit test is robust. Fourthly, the case of correctly specified alternatives is considered and low power of the ESTAR but not for the MSAR unit root test is observed. Fifthly, an empirical application to real exchange rates suggests that they may indeed be explained by Markov Switching dynamics rather than ESTAR.

KW - ESTAR

KW - Markov Switching

KW - PPP

KW - Real exchange rates

KW - Unit root test

UR - http://www.scopus.com/inward/record.url?scp=84865820132&partnerID=8YFLogxK

U2 - 10.1007/s00181-010-0431-2

DO - 10.1007/s00181-010-0431-2

M3 - Article

AN - SCOPUS:84865820132

VL - 43

SP - 457

EP - 474

JO - Empirical economics

JF - Empirical economics

SN - 0377-7332

IS - 2

ER -