This thesis investigates the intraday lead-lag relationship between returns on KOSDAQ50 futures and returns on its cash index using the 5-minute price date for the period of 2001.01.30- 2001.09.13. Using the cointegration analysis and error correction model, lead from the cash to the futures market has been confirmed. These results are inconsistent with the previous studies, such as Stoll and Whaley(1990) and Eun and Chang(1998). This result may be interpreted as inefficiency in the future market caused by lack of trading volume.
First, KOSDAQ 50 cash index tended to lead the future prices by thirty minutes and the lead effect of future prices was disappeared in five minutes. The magnitude of the coefficient on the futures first lagged variable is insignificant.
Second, through the comparison between the non filtered error correction model and the ARMA filtered error correction model, the lag of significant coefficient in futures become shorter. It explains that infrequent trading in cash market is one of reason for lead in future market.
Third, the additional sub period analysis has been performed to discover improvement of future market efficiency. The magnitude of the coefficient on the futures lagged variables has been improved with trading volume of future market increasing. Increasing trading volume of future market involve increase of speculative trading and arbitrage trading.
