Learning by Doing in Crowdlending Auctions

Abstract

We construct a novel taxonomy of bidding strategies in Dutch auctions, associated with five types of investors in those auctions: angel bidders, irrational bidders, experts, followers, and snipers, the last three types being considered as strategic. By applying this classification to peer-to-business crowdlending data, we show that: (1) retail investors become more strategic with experience; (2) winning auctions makes investors more likely to be snipers while losing auctions makes them more likely to behave as experts or followers; (3) among experts, those who are early members of the platform are found to be informed about the quality of auctioned loans.

Keywords: bidding strategies, crowdlending, learning effects

JEL Classification: G11, G23