(1999) study that discovered a number of cointegration relations in their sample. By focusing on
underlying dynamics of the relationship as reflected in the recursive cointegration test, we infer
that Olienyk et. al (1999) findings could have been driven by the choice of their sample period.
The third period ranges from October 2000 to March 2001 and indicates disruption of long-run
equilibrium with the minimum number of cointegration relations falling to 1. Finally the fourth
period ranges from April 2001 to January 2005 and reflects increasing convergence with the
minimum number of cointegration relations rising to 2.
Overall, our graphical plots based on recursive cointegration test suggest a trend towards
increasing integration since 2001. In an earlier study, Rangvid (2001) analyzed dynamics of
integration between the major European equity markets using quarterly data on IFS indices from
1960 to 1999 and found single cointegration relationship. Having applied the recursive
cointegration test of Hansen and Johansen, Rangvid also pointed to increasing degree of
European financial markets integration as reflected in the upward trend of recursive lambda trace
statistics, though the convergence did not occur until 1982.
6.3. Dynamic Conditional Correlations (DCC-GARCH) Analysis Results
Dynamic conditional correlations between SPDR and iShares returns, calculated as
described in Section 4.2.3, are presented in Figure 3. The graph shows the evolution of the
dynamic conditional correlations over time. A number of regimes can be distinguished.
In the first regime, prior to 1997, conditional correlations are found to be generally
declining. The second regime that coincides with the financial turmoil of 1997-1999 is
characterized by a drastic increase in the value of the conditional correlations. For example, in
case of German iShares, conditional correlations with SPDR exhibit an increase from 0.25 to 0.7
during this period. This finding is in line with the classical work of Longin and Solnik (1995)
21