Abstract:
This thesis proposes a trading strategy which trades based on the observation and prediction of abnormal volume events of stocks listed on the Stock Exchange of Thailand (SET) during July 2010 to June 2016. This research found that a positive excess return follows an abnormal volume event defined by the abnormally-high standardized volume and standardized directional volume. To confirm that such events are exploitable, a strategy that trades on those events is tested and found that they generate positive alphas even after including commission fees. Previous work has shown that typically an abnormal volume event is accompanied by a substantial excess return on the same day. Thus, this thesis further improved the strategy by attempting to capture the excess returns on the same day as abnormal volume events. An algorithm capable of predicting those events with high precision is developed and integrated into the strategy, enabling trade initiation before the end of the day. A portfolio simulation on out-of-sample data shows that the intra-day strategy generates incremental excess returns.
