Current Issue: Volume 12, Issue 3 (2008)
Regime-Switching Models in Economics and Finance

Introduction

Guest Editors: Gary Koop (Strathclyde University); Costas Milas (Keele University); and Denise Osborn (Manchester University). Nonlinear time series models in general, and regime-switching models in particular, have increased our understanding of many issues in economics and finance. This special issue collects papers presented at the first conference in the Economic and Social Research Council (ESRC) Seminar Series on “Nonlinear Economics and Finance Research Community”. This took place at Keele University (UK) on February 2, 2007 and was hosted by Christopher Martin (Brunel University), Costas Milas (Keele University) and Theodore Panagiotidis (University of Macedonia), with funding by the ESRC under grant RES-451-25-4260. The aim of this Seminar Series is to bring together researchers working on nonlinear topics. The papers included in this special issue cover a range of topics. The paper by Granger (the keynote speaker at the conference) explains why models with time-varying parameters may be potentially more useful than non-linear models. Markov-switching models provide one approach to capturing time-variation through a non-linear specification. In this context, the paper by Rothman introduces a Markov Chain test and identifies asymmetries in aggregate and disaggregated US unemployment rate data. The paper by Kim and Kim uses a Markov-switching model to investigate changes in the US New Keynesian Phillips Curve. Nankervis, Sajjad and Coakley also adopt a Markov switching approach to explore the ability of ARCH-type models to forecast value at risk using data on US and UK stock returns. On the other hand, the paper by Juvenal and Taylor uses a threshold approach to identify nonlinear mean reversion in deviations from the law of one price using data for nineteen sectors in ten European countries. Some contributions focus on theoretical or methodological issues. The paper by Hu and Shin introduces specification tests against Markov-Switching GARCH models and uses these tests in US, UK and Japanese stock return data. Hultblad and Karlsson develop a new Bayesian approach to simultaneously model lag length and structural change and use their approach in an application involving the US real interest rate. Finally, the paper by Harvey, Leybourne and Xiao introduces a new powerful test against nonlinearity when the order of integration of the variable is unknown.

Articles