We investigate a unique dataset that includes detailed bidding information of 783 Chinese IPOs. First, we find that institutional investors who participate in multiple IPOs on a single day suffer from limited attention and submit less precise bids. Second, we form several proxies for investor attention capacity and show that investors with larger attention capacities are less affected by the inattention bias. Furthermore, we show that investors are more affected by the inattention bias in IPOs that are more difficult to price. Moreover, we find that the allocation rule change from pro-rata to lottery leads to a reduction in investors' multiple participations, and improves their bidding precision. Third, we show that investors are more likely to submit lower bids when participating in multiple IPOs. Consequently, the investors affected by the inattention bias are allocated fewer shares and earn less profit. Finally, we find that aggregate investor inattention at the IPO level significantly reduces the IPO's pricing efficiency.
