Acceptable Risk in a Portfolio Analysis

A social network has been used to simulate how agents of different levels of risk aversion under different circumstances behave in financial markets when deciding between risk-free and a risky asset. This is done by a discrete time version evolutionary game of risk-loving and risk-averse agents. The evolutionary process takes place on a social network through which investors acquire information they need to choose the strategy. A significant feature of the paper is that first-order stochastic dominance is a key determinant of the decision-making, while second-order stochastic dominance is not, with the level of omniscience and preferences of agents also having a significant role. Under most of the circumstances, pure risk-aversion turns out to be dominated strategy, while pure risk-taking “almost” dominant