Determinants of Amount Raised in Equity Crowdfunding Campaigns: An Application of Truncated Regression

Abstract

Secondary-market data on 1189 crowdfunding campaigns, extracted from the leading UK equity crowdfunding platform Seedrs, is used to identify the determinants of amount invested. A key feature of the dataset is that only successful campaigns are observed. For this reason, we depart from previous literature by applying the truncated regression estimator instead of ordinary least squares regression. We cluster at the company level in order to allow for multiple campaigns per company. The importance of allowing for truncation is confirmed using the Hausman test. Moreover, allowing for truncation alters the conclusions arising from previous literature concerning the effects of standard variables, including target amount, amount of equity offered, and number of team members participating in the venture. Quadratic terms are introduced in order to identify “sweet spots.” Textual dummies are included in order to identify words that enhance investor appeal. The Linktest misspecification testing procedure confirms the validity of our final model.