Knowledge, Preferences and Shocks in Portfolio Analysis

We simulate social network games of a portfolio selection of agents to analyze how knowledge, preferences of agents and their level of omniscience affect their decision-making. The key feature of the paper is that preferences and the level of omniscience of agents very much determine the ways agents make their decision. By introducing a one-time shock, we found that its efficiency depends on the level of omniscience of agents, with much stronger efficiency under omniscient agents