Presentation av mastersarbete. Titeln på examensarbetet är "Conflict of Interests in an Artificial Stock Market: How Irrational Traders Interact with Rational Investors".
This thesis presents results from the trading interactions of various types of agents based on the Santa Fe Institute artificial stock market (SFI-ASM). The primary agents learn to optimize their expected returns based on the information they obtain from the market. Two other types of agents trade along with the primary agents in our ASM, underinformed investors, and regression traders, and analyze the resulting market dynamics. Underinformed traders receive noisy fundamental information (related to the intrinsic asset value) during their forecast update and prediction accuracy assessment. Regression traders look into actual or filtered price values and estimate their forecasts based on previous prices. The prime motivation behind this survey is investigating how the inclusion of irrational traders influences the market and other already established traders. We find that being underinformed is not strictly detrimental in the context of this model. Our results indicate that when the number of irrational traders increases, their profits are more likely to decrease, and rational agents can exploit them more readily. However, when there are a few irrational traders, they can gain more returns. We also find that the addition of trend followers results in a bullish market early after the involvement of the said traders. Our experiments suggest that in order for the agents to take advantage of a certain irrational strategy, the number of practitioners of that strategy should be larger than a threshold. This last finding sheds new light on the ramifications and premises of the efficient market hypothesis.
Zoom-länk: https://chalmers.zoom.us/j/68930442968