For MENA markets, there were no connections with Latin American markets and only weak connections with markets in the other two regions. The Asia-India correlation was found to be strongly influenced by market characteristics and the GFC, while the Asia-China correlation was unaffected by any factor considered for the monthly correlation. Several previous studies investigate stock market integration in MENA using the DCC-GARCH approach, including those of Maghyereh et al. 2015) consider daily stock prices in US and MENA markets for 2005–2013 to investigate stock market integration before and after the GFC.
They also examine the reasons for the identified relationships and find that increasing stock market integration in these countries was not due to any macroeconomic or monetary convergence, but rather due to developments in the financial sector. Panda and Nanda (2018), Hwang (2014) and Lahrech and Sylwester (2011) examine stock market integration in the Latin American region using the same DCC-GARCH approach and draw on weekly data from 1995 to 2015 and find that co-movements was higher towards the end of the trial period than in the early phase. The first step requires estimating the variances via a univariate ARMA-GARCH specification, and the second step requires estimating the parameters that capture the dynamic nature of the correlations.
That is, after fitting the univariate ARMA-GARCH models, in the second step of the DCC model, standardized residuals are used to estimate the dynamic correlations. This study uses daily MSCI-based stock market indices of the emerging markets of Asia, CEE, MENA and Latin America over the period 3 January 2002 to 11 November 2016. The average of the MSCI returns is on average the strongest for the Latin American region, following through Asia, CEE and MENA.
In general, the ARMA (1,1) GARCH (1,1) procedure is found to be the best candidate to estimate the first step of the DCC-GARCH approach based on the Akaike Akaike information criterion and the Schwarz criterion.
EMPIRICAL RESULTS
First, previous studies using data taken from years before 2012 note that the time-varying correlations were highest during the GFC and higher in the post-GFC period than the pre-GFC period. The DCC link between each of the Latin American nations and other emerging countries is the weakest, with DCCs falling in the -6% to 6% range (Table 5). Third, the DCCs between South Africa and Malaysia averaged 63% for the period, making the two markets the most connected of all markets in the sample (Tables 2 and 4).
The DCCs between Lithuania and Mexico were -6%, making these two the least connected markets in the sample (Tables 3 and 5). For some Asian and CEE markets with Oman, Morocco, Colombia, Mexico, Pakistan and Kuwait, we find that the GFC was strongly associated with a decline in DCC. That is, DCCs increased during the GFC and decreased during the non-GFC period.
However, we see that India's DCCs with markets in selected countries, such as Thailand, Tunisia, Hungary, Poland, South Africa, Croatia, Romania and Lebanon, were higher in the post-GFC period (Table S1). The DCCs between Indonesia and MENA were higher during the post-GFC period compared to the GFC or pre-GFC period (Table S1). The post-GFC DCCs for Pakistan were higher than the DCCs during the GFC or the pre-GFC period for all the above countries except India, which had higher DCCs in the GFC period.
Similarly, Thailand had DCCs with CEE (16–34%) and MENA (8–35%), with evidence of increased interdependence with Kenya, South Africa and Croatia in the post-GFC period. This period coincided with a significant decline in CEE countries' DCCs with other emerging markets. However, with Romania, Oman and Pakistan, Estonia's DCCs were stronger in the post-GFC period.
Hungary's time-varying correlations, on average, were in the range of 25-55% with other CEE countries, 7-49% with Asian countries, and 5-44% with MENA countries. The post-GFC period saw a decline in DCCs among Hungary and all other nations in the sample except Romania, India, Pakistan, Sri Lanka, Kenya, Morocco and Tunisia. Conditional Dynamic Correlations: CEE Region vs Other Emerging Markets This table presents the pairwise DCCs over the period 2002-2016 for each CEE market in column 1 against the other emerging markets listed in the corresponding rows .
In the post-GFC period, 9 out of 15 countries saw an increase in the number of DCCs compared to the GFC period: Romania, India, Malaysia, Thailand, Korea, Sri Lanka, Kenya, Greece and Poland. Post-GFC DCCs were larger than those during the GFC period for all countries except Korea, Kenya, Tunisia and Greece (Table S2).
MENA
CONCLUSION
This study sought to determine whether a trader holding a share in an MSCI-based stock market index in his home country – in Asia, CEE, Latin America or MENA – has short-term diversification opportunities in any of these four new regions. Some of the common themes that emerge from this research in relation to each region are as follows. The DCCs between the CEE markets, except for Greece, increased during the ESDC period, indicating contagion.
In the case of MENA markets, there were no connections to Latin American markets and only weak connections to the markets in the other two regions. However, the South African stock market was an exception, which proved to be highly integrated with Asian markets. The Latin American countries were found to have significant short-term connections to the MENA countries, but very low connections to other markets, both in Latin America and CEE countries, making emerging and frontier markets lucrative for short-term diversification.
In terms of diversification opportunities, traders with home-biased portfolio in the Latin American region have the most opportunities to diversify their portfolio in the emerging/frontier markets in their region and other two regions. This is for those traders who live in one of the four regions and have a home biased investment. When we considered CEE's DCCs before, during and after the ESDC, we found that the DCCs between Greece and other emerging/frontier markets generally fell, but that other CEE nations increased their DCCs with the emerging/frontier markets of the other three regions.
The Israeli-Hezbollah war and the global financial crisis in the Middle East and North African capital markets. Stock market contagion from Western Europe to Central and Eastern Europe during the crisis years 2008-2012. Local Does as Local is: The Informative Content of the Geography of Collective Investments of Individual Investors.
Integrating Global Leaders and Emerging Powers into the Malaysian Stock Market: A DCC-MGARCH Approach. Volatility spillovers due to differences in the degree of market integration: Evidence from selected Asian and Eastern European stock markets. The Impact of Terrorism Risk on Stock Market Integration: Evidence from Eight OECD Countries, International Review of Financial Analysis.
