Working Paper: CEPR ID: DP5352
Authors: Peter C. Schotman; Anna Zalewska
Abstract: The paper contributes to the literature on integration of stock markets by addressing the issue of non-synchronous trading. We argue that controlling for time differences in trading hours of stock markets is important and show that time-adjustment improves estimates of market integration. We also show that using weekly frequency does not sidestep the consequences of the time-match problem but leads to significant loss of information. We show that the nature of integration of stock exchanges operating in the Czech Republic, Hungary, and Poland with the stock markets of Germany, UK and US in the period 1994-2004 is very dynamic. Finally, the study shows that the autocorrelation of returns on the main market indexes of the emerging markets have declined over time.
Keywords: Emerging Markets; Market Efficiency; Market Integration; Kalman Filter; Nonsynchronous Trading
JEL Codes: G14; G15
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Controlling for time differences in trading hours (C41) | Accurate estimation of market integration (F15) |
| Time-adjustment (C22) | Stronger correlations between CEE markets and developed markets (F36) |
| Nonsynchronous trading (G19) | Misinterpretations of market behavior (G41) |
| Time-matched daily data (C22) | More dynamic integration process (C69) |
| Decline in autocorrelation of returns on CEE market indexes (C22) | Shift towards increased market efficiency (G14) |
| Responsiveness of CEE markets to external shocks (F44) | Understanding of market predictability and efficiency (G14) |
| Time-aligned data (Y10) | Avoiding biases in estimating market integration (F15) |
| Non-matched data (Y10) | Failure to yield statistically significant results (C90) |