How Much Artificial Intelligence Do Robo-Advisors Really Use?

Abstract

To understand how and why robo-advisors implement artificial intelligence (AI), we use evidence from structured interviews with C-level executives from a large group of robo-advisors. We find that robo-advisors see more potential in sales and client directed areas than in portfolio management. This finding can be explained by the business side of robo-advisory. Actual implementation of AI, on the other hand, seems to be driven by workability and not by potential. The level of engagement in AI is higher for companies that focus on portfolio management; a focus on other areas does not significantly influence engagement in AI. Portfolio management is also the area where AI initiatives are most advanced. Even though robo-advisors attach moderate to substantial importance to AI and concrete measures toward AI implementation are noticeable rather broadly, so far only a minority actually use AI. The main obstacles for AI implementation as perceived by the management are budget restrictions, unclear value-added, lack of experts, and regulatory hurdles. Furthermore, the noticeable hindrance for AI seems to be largely market-driven hesitancy as opposed to concrete difficulties.