Has entry to the European Union altered the dynamic links of stock returns for the emerging markets?

Abstract

This article investigates the impact of the entry to the European Union (EU) on the dynamic links between the stock market indices of Czech Republic, Hungary, Poland and Slovakia vs. those of the euro-zone by utilizing the international version of the feedback-trading model. Prior to entry, there was evidence of feedback trading with the euro-zone, however, this disappeared in the post-entry period with the exception of Slovakia. Evidence appears to demonstrate the emergence of financial integration of these transition economies within the EU.

Acknowledgements

The author thanks Christopher Green and Theodore Panagiotiolis for valuable comments, as well as the participants at the International conference on European Labour Markets, Trade and Financial Flows and Economic Growth, held in Crete, Greece, May 2006.

Notes

¹ The smooth transition trend model can indicate the speed at which a market is becoming integrated (Chelley-Steeley, 2005).

² This is empirically evidenced by Sentana and Wadhwani (1992) for UK stock market. The authors find that, during high volatility periods, daily stock returns tend to be negatively autocorrelated.

³ There has been work to identify a number of key macroeconomic variables (e.g. output, inflation and interest rates) in terms of stock market co-movements (e.g. Dickinson, 2000). The specification of such variables into the model is beyond the scope of the current investigation.

⁴ There is evidence that stock returns across national markets exhibit time-varying correlations (e.g. Tse, 2000). Empirically, the presence of a strong influence of US stock returns on other markets is observed. Therefore, although the time zone is different, the US market is included.

⁵ The inclusion of the currency return is consistent with the data in that, the US or euro-zone stock returns can be converted into the value of returns from the point of view of the investors in the new EU member states.

⁶ In Slovakia, the null is rejected in the Trace test, but not rejected by the Maximum test.

⁷ See Azman-Saini et al . (2002) and Piesse and Hearn (2002).

⁸ As to Poland, Hungary and Slovakia, cointegration results were ambiguous and also the coefficient on the error correction term had a wrong sign. Therefore, we do not pursue the error correction model for these countries.

⁹ Note that the European markets are about to close when the US market opens.

¹⁰ It is argued that possible factors which contribute to the phenomenon, include portfolio insurance strategies, the extensive use of stop-loss orders and margin trading.

¹¹ Moreover, EMU implies monetary policy convergence suggesting a fall in the variance of interest rates and inflation rates for these economies. The decline in volatility of European stock markets may be attributed to the stabilization of fundamentals.

¹² In the empirical work of Westermann (2004), the stock returns for the euro-zone are modelled with the daily data for the sub-sample period of 1998 (before the euro) and 1999 (after the euro) separately. The author finds the significant positive feedback trading across the euro-zone countries during 1998, however, it has disappeared after the introduction of the euro.