Detecting jumps and regime switches in international stock markets returns

Abstract

This article explores seven international stock markets (DJIA, Euro STOXX 600, Russell 2000, Nikkei, NASDAQ, FTSE and Global Dow) in the quest for jumps and regime switches. The methodological framework borrows from the Markov-switching approach and the stochastic modelling literature based on Lévy processes. The econometric procedure is detailed in a two-step fashion. The data set covers the period from June 2004 to July 2014. The main results uncover changing market dynamics according to economic and/or financial phenomena (e.g., economic crises/growth, news events) with the occurrence of several episodes characterized by a high jump intensity. We advocate the use of such a jump-robust model modulated by a Markov chain to further study the dependence structure of financial time series.

Keywords:

• Lévy jumps
• Markov switching
• equity markets
• Nasdaq
• Euro STOXX
• FTSE

JEL Classification:

• C32
• G15
• E44

Notes

¹ It is well known that testing for the number of regimes in a Markov chain is a hard problem to tackle, which we leave for further research.