SUSTAINABLE EQUILIBRIUM IN A STOCK MARKET: AGENT-BASED MODELING WITH EVOLUTIONARY GAME THEORY APPLIED TO TRADERS

Tugba Karabiyik
Sam Houston State University

Orhan Akal
Florida State University

Elvan Aktas
Valdosta State University

ABSTRACT
This study employs an Agent-Based Model with Evolutionary Game Theory. First, we utilize a stock market simulation with four heterogeneous trader types: Momentum, contrarian, long term and speculative. They have deterministic decision rules, and they are given realistic trading conditions such as wealth constraints and learning behaviors. Their interactions are applied to a simulated stock market where we were able to replicate the quasi-random dynamic behavior of an actual stock market. Each trader also has the ability to change its trader type based on its past trading performance and its competitors past performance. In the long run equilibrium, long term traders dominate the stock market; the number of momentum and contrarian traders remain relatively low. In terms of relative total wealth, however, speculators hold almost half of the entire wealth that is invested in the stock market simulation. The results are very realistic compared to the long-term sustainable equilibrium in a continuous price-discovery process, such as the stock market. Secondly, we focus on understanding the behavior of the stock market traders by utilizing an evolutionary game theory model. This study is the first in literature to employ such theoretical analysis with four types of traders. The authors allow each trader type to have its own strategy to make trading decisions in the stock market to maximize wealth. Each trader tries to maximize its payoff by changing the trading strategies with the consideration of learning and wealth constraints. However, each trading strategy will incur two types of costs: time value of money and transaction costs. The authors predict a variety of realistic trading strategies that have been documented in real-life stock market equilibrium.
Keywords: Stock market simulation, game theory, agent-based modeling, market equilibrium