Abstract: From the investors’ perspective, crowdfunding is fundamentally about selecting ideas by aggregating individual judgments into a group decision under uncertainty. Compared with traditional investment activities, crowd-based investment is characterized by being grass-rooted, democratic and diversified. Based on these characteristics, this paper further divides the crowdfunders’ decision-making process into three stages: independent decision-making, belief aggregation and consensus emergence. A computational experimental method is introduced to simulate the crowdfunding practices. The results show that the cognition heterogeneity of investors and the formation mechanism of group consensus are the key factors determining the performance of crowd-investing. Specifically speaking, (1) the group diversity can act as alternative to ability. When the number of crowdfunders is large, group diversity is even more important than ability. (2) The financing threshold is conducive to narrowing down the choices and screening the good projects, and the higher the financing threshold (i.e. the consensus level), the better the screening effect. Therefore, from the perspective of investors, AON is always better than KIA. (3) Increasing difficulty of assessment may weaken the effect of cognition heterogeneity.

Key words: crowdfunding, individual cognition, group consensus, investment performance, computational experiment