| Gervais et al. (2001) present the visibility hypothesis---the concept that extreme trading activity contains information about the future evolution of stock prices. The study argues that this high-volume return premium is consistent with the idea that shocks in the trading activity of a stock affect its visibility, and in turn the subsequent demand and price for that stock. The results imply that trading volume movement may foresee future price changes.; This study will utilize the learning algorithms of a neural network to empirically investigate the predictability of trading volume changes. According to Fama's (1970) definition, the term "predictabilities" refers to how the possibilities offered by the information transference of international markets in different time zones may provide superior profits from equity trading. General research questions being examined are stated as follows: (1) The first research question applies the trading signals produced by the neural trading systems to trade on the closing position of the TSM, if the "Signal V" based return outperforms the buy-and-hold based return, it may support the existence of Gervais et al.'s (2001) visibility hypothesis. (2) The signals will be further utilized to simulate a "no position holding strategy", and compared to the previous simulation results. This experiment will simulate the scalpers' (day-traders') trading strategies, and may compare the predictability of trading volume in day-trading activities. The empirical results may be utilized to explain the phenomena of relatively higher turnover. |
