Investor education innovation based on generative artificial intelligence : a randomized survey experiment on disposition effect

Abstract

Currently, investor education faces many challenges. The traditional model is costly and limited to basic financial knowledge. In response to this problem, this study innovatively introduces generative artificial intelligence into the investor education scenario. Through a randomized survey experiment, taking the disposition effect as an example, it compares the educational effects of three different forms of investor education based on generative artificial intelligence: comics, stories, and academic content. The study finds that comic content can more effectively reduce investors’ disposition effects compared to academic content. The mechanism lies in enhancing awareness of behavioral biases, reducing related negative emotions, and improving cognitive levels, as well as the heterogeneity of the population and differences in various aspects of different contents. This research provides empirical evidence for the application of generative AI in the field of investor education and is of great significance for improving educational methods, promoting financial market stability, and promoting the development of inclusive finance.