Journal on Policy & Complex Systems Vol. 2, Issue 2, Fall 2015 - Page 51

Journal on Policy and Complex Systems - Fall 2015, Volume 2, Number 2 Enhancing Stock Investment Returns with Learning Aggressiveness and Trust Metrics Zheyuan Su and Mirsad Hadzikadic Complex Systems Institute, College of Computing and Informatics, University of North Carolina at Charlotte, Charlotte, NC 28223, USA Email: and Abstract: Trust metrics and learning can be extremely useful in enhancing the overall investment returns. A trust metric is an indication of the degree to which one social actor trusts another, while aggressiveness in learning determines the degree to which one trader decides to mimic another. This paper introduces an agent-based model for finding the optimal level of aggressiveness in learning and the optimal degree of trust in order to optimize the stock-trading returns. The system has been evaluated in the context of the Bank of America stock and S&P 500 performance in the period of 1987–2014. The model significantly outperformed the buy-and-hold strategy on both S&P 500 and the Bank of America stock. In addition, the model can provide relevant information to policy maker regarding interest rate setting and expected investor behaviors. Keywords: complex adaptive systems, investment, trust metrics, learning aggressiveness 1. Introduction F inding best stock investment strategies requires not only rational trading rules practices but also the faith that market information is reliable. A trust metric is an indication of the degree to which one social actor trusts another. It is hard to pick the right timing for stock selection or to take the correct position in the stock market, because stock intrinsic values are affected by both endogenous and exogenous phenomena. Some investors actually make profit by taking advantage of the asymmetric property of information, which is what happens when one party in a transaction has better information than the other party does. However, actions taken based on asymmetric information are reflected in the 49 doi: 10.18278/jpcs.2.2.4