Fostering the adoption of robo-advisors: A 3-weeks online stock-trading experiment

We consider how to increase the take-up of robo-advisors to help investors cope with the disposition effect. In a 3-weeks online stock-trading experiment, participants traded freely in week 1, were assisted by trading algorithms in week 2, and chose whether to be assisted in week 3. Passive algorithms prevented trading, active ones traded according to Bayesian rules. Participants could override algorithm choices in some treatments. Only a minority adopted robo-advisors in week 3, with the worst performers being the least likely to do so. Algorithm aversion was reduced if the algorithm traded actively and could be overridden.