Crowdfunding with Sequential Moves and Costly Common Value Signals

• Authors: Matthew Ellman and Sjaak Hurkens
• BSE Working Paper: 1548 | January 2026
• Keywords: information acquisition, crowdfunding, perfect bayesian equilibrium rationalizability
• JEL codes: C72, C73, D82, D83

Cong and Xiao (2024) show that All-or-Nothing crowdfunding, modelled as a sequential game with observational learning and free private information, achieves efficiency and extracts all surplus in the asymptotic limit as the crowd size grows. Ellman and Hurkens (2025a) model investment-based crowdfunding as a simultaneous move game with costly information acquisition. They establish a single upper bound on per capita welfare and profit in symmetric equilibria that is shown to be tight in asymptotically large crowds. This paper shows that with costly information, sequentiality allows for strictly higher welfare but not for higher profit.