Effect of Information Exchange in a Social Network on Investment

Abstract

Herd effect in a multi-agent model with a static social network is investigated. The agents are playing the Parrondo’s game which can be considered as an investment game into two slot machines, C and D, so that playing continuously on one slot machine will lose, but by suitable switching the play on these two slot machines a player can win in the long run. This strange effect has its origin in the non-equilibrium physics of Brownian ratchets and can be analysed using Markov chain. The impact of information exchange on the collective behaviour of the players is investigated in a social network, with the players adopting one of two strategies: ‘Follow the winner’ or ‘Avoid the loser’. Players using either strategy alone will lead to loss for the entire population. This herd effect is observed numerically and explained using Markov chain. For the ring type social network, the population can achieve positive gain with two protocols for information exchange when the players communicate with their nearest neighbours. The first protocol is to randomly mix the game strategy ‘Follow the winner’ with ‘Avoid the loser’. The second protocol is to give each player a pre-set probability to switch between ‘Follow the winner’ and ‘Avoid the loser’. We provide a heuristic explanation for the difference in gain of these two protocols based on the probability distribution of the resident time for the player in his selected strategy. We also discuss the evolution of their wealth distributions.