Header menu link for other important links
X
Testing Non-linearity in Emerging and Developed Markets
, Basabi Bhattacharya, A Roy Chowdhury
Published in Springer India
2016
Pages: 437 - 465
Abstract
Behaviour of non-linearity in stock markets of developed countries has been empirically evidenced for quite some time. Some few studies examine non-linearity of emerging stock markets as well. A comprehensive overview of stock market behaviour in terms of non-linearity is undertaken in this study where several non-linearity test approaches have been applied on both the developed and emerging country groups. For the purpose we have selected seven stock exchanges from developed markets (UK, Germany, Australia, New Zealand, Hong Kong, France and Japan) and seven from emerging markets (Hungary, Chile, Mexico, Malaysia, Taiwan, China and India). Taking their daily close data covering a period from January 2005 to July 2011, we have investigated the presence of nonlinearity in the data sets and tried to find out whether there is any difference amongst them in this regard. We have first used the test method developed by Brock, Dechert and Scheinkman (BDS) and examine the non-linearity feature in each of the time series. To reinforce our findings, we have conducted the White's Neural Network tests on the same data set. Additionally, we have also performed the Keenan's test for non-linearity. Another popular non-linear test is the Hinich bi-spectrum test, which involves estimating the bi-spectrum of the observed time series. We have further used this test to find out whether it detects non-linearity in these time series. Another linearity test for time series was introduced based on concepts from the theory of neural networks. Ter"asvirta et al. developed its power fully. We have applied this Ter"asvirta Neural Network test as a final reinforcement of our findings.
About the journal
JournalData powered by TypesetInternational Trade and International Finance
PublisherData powered by TypesetSpringer India
Open AccessNo