A Report of U-Mart Experiments by Human Agents

Regarding whether the board information was open or not, the null hypothesis was not rejected. There is no significant difference in the rate of contracted volume per volume of order. Although statistical theory provides a rationale for rejecting the null hypothesis, it provides no formal rationale for accepting it. Although the result of the experiments was not statistically significant, there may be another possibility that can be concluded from further experimentation. The Gaussian distributed price change was not ascertained by the experiments. This is the first series of U-Mart experiments that were performed systematically by more than 20 human trader agents and machine agents. However, the number of experimental data is not large, and the results obtained are restrictive. Further experiments are needed to improve our knowledge about the financial market.