Boys will be boys, but robots can help them: Gender, algorithmic compliance, and portfolio performance

Abstract

We document a gender gap in the adoption of AI algorithmic advice. Using data from a Korean robo-advisory platform where users must manually approve trades, we find men are significantly more likely to comply with rebalancing proposals promptly than women. We trace this gap to execution frictions. Men execute trades rapidly and capture the algorithm's long-term annual performance premium, earning a 2.4% return to compliance. Women delay approval, creating an execution wedge that renders the advice statistically unprofitable for them. Our findings suggest that even in automated investing, human execution mechanics remain a critical determinant of portfolio performance.