The Links Between Central, East European and Western Security Markets

Over the last few years, Central and East European economies have become more integrated with the West European economy. An important step in this development was the admission of eight East European countries to the EU in 2004. As the security markets in Europe integrate, an important issue is whether West European investors can still gain diversification benefits from investments in the stock and bond markets of Central and Eastern Europe? In this chapter we try to answer this question and we focus on how the West, Central and East European security markets are linked with each other, both in the short term and in the long term. In contrast to earlier research we do not analyze the stock markets or the government bond markets in these regions separately, but we analyze both markets and their interaction. We think it is important to analyze bond and stock markets at the same time, as most investors tend to hold a diversified stock-bond portfolio. Furthermore, we include the volatility of these markets in our analyses. We calculate correlations between the security market returns to measure short-term links. We also apply the Granger (1969) causality test, which includes lagged returns. To assess the potential long-term links between security markets, we use the well-known Johansen (1991) cointegration test and the Granger (1969) causality test. Our data set consists of the stock markets and the euro-denominated government bond markets of the region Western Europe, three Central European countries, namely the Czech Republic, Hungary and Poland, and two East European countries, Russia and Turkey. The stock market total returns cover the time period February 1997 to November 2004, on a weekly basis. The euro-denominated government bond market total returns cover the shorter time period October 2001 to November 2004, again on a weekly basis (data for the Czech Republic is not available). One of our main findings is that there are clear links between stock markets of Western Europe, and Central and Eastern Europe, but still there is ample opportunity for risk reduction through portfolio diversification. The maximum short-term return correlation is 51 percent, between Poland and Western Europe. We find no evidence of a long-term cointegration relationship between the stock markets. Moreover, the Granger (1969) tests show that no security market dominates all other security markets. For the euro-denominated government bond markets, the picture is completely different: The short-term return correlation between Western Europe and two Central European countries, Poland and Hungary, is 90 percent or more. Diversification of euro-denominated government bond portfolios is still possible through investments in the Russian and Turkish markets though, as the short-term correlation of these below investment grade markets with Western Europe is 27 percent or less. Granger (1969) tests show that no bond market leads all other bond markets. Finally, we find the best opportunities for diversification in stock-bond portfolios: The short-term correlation between Western European stocks and bonds is -55 percent in our sample period from October 2001 to November 2004. We find similar negative correlations for the stock-bond relationship in Hungary and Poland.