From Cognitive Bias to Market Dynamics: Teaching Behavioral Finance with an LLM-Based Agentic Market Simulation 

Chu Yinxiao ¹  Huang Xiaoyu ²

1 Huazhong University of Science and Techology, Wuhan, Huibei China，430074；

2 Wuhan University of Technology, Wuhan, Huibei China，430070；

Abstract：Behavioral finance is difficult to teach because students must connect individual bias, trading behavior, social interaction, and market-level outcomes within the same explanatory frame. Conventional teaching often presents these layers separately, leaving the field conceptually clear but pedagogically fragmented. This paper develops a design-oriented framework for teaching behavioral finance through an LLM-based market simulation. Using TwinMarket as an illustrative case, it shows how such a platform can make behavioral processes visible for analysis, comparison, and reflection. The discussion explains why behavioral finance benefits from scenario-based learning, what distinguishes TwinMarket from more familiar AI teaching tools, and how simulation-based instruction can be organized, assessed, and governed in practice. The framework emphasizes staged task design, theory-guided debriefing, process-oriented assessment, and explicit governance boundaries. The result is an implementable teaching design centered on instructional logic, classroom organization, and responsible use of simulation in higher education.

Keywords: behavioral finance; large language models; market simulation; instructional design; TwinMarket; higher education

References

[1] Baker, M., & Wurgler, J. (2007). Investor sentiment in the stock market. Journal of Economic Perspectives, 21(2), 129-152. https://doi.org/10.1257/jep.21.2.129

[2] Barber, B. M., & Odean, T. (2000). Trading is hazardous to your wealth: The common stock investment performance of individual investors. The Journal of Finance, 55(2), 773-806. https://doi.org/10.1111/0022-1082.00226

[3] De Long, J. B., Shleifer, A., Summers, L. H., & Waldmann, R. J. (1990). Noise trader risk in financial markets. Journal of Political Economy, 98(4), 703-738. https://doi.org/10.1086/261703

[4] Flavell, J. H. (1979). Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist, 34(10), 906-911. https://doi.org/10.1037/0003-066X.34.10.906

[5] Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263-291. https://doi.org/10.2307/1914185

[6] Kasneci, E., Sessler, K., Kuchenmann, S., Bannert, M., Dementieva, D., Fischer, F., Gasser, U., Groh, G., Gunnemann, S., Hullermeier, E., Krusche, S., Kutyniok, G., Michaeli, T., Nerdel, C., Pfeiffer, F., Poquet, O., Sailer, M., Schmidt, A., Seidel, T., … Kasneci, G. (2023). ChatGPT for good? On opportunities and challenges of large language models for education. Learning and Individual Differences, 103, Article 102274. https://doi.org/10.1016/j.lindif.2023.102274

[7] Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and development. Prentice-Hall.

[8] Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge University Press.

[9] Park, J. S., O’Brien, J. C., Cai, C. J., Morris, M. R., Liang, P., & Bernstein, M. S. (2023). Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology (pp. 1-22). ACM. https://doi.org/10.1145/3586183.3606763

[10] Shleifer, A., & Vishny, R. W. (1997). The limits of arbitrage. The Journal of Finance, 52(1), 35-55. https://doi.org/10.1111/j.1540-6261.1997.tb03807.x

[11] Shulman, L. S. (1986). Those who understand: Knowledge growth in teaching. Educational Researcher, 15(2), 4-14. https://doi.org/10.3102/0013189X015002004

[12] Tlili, A., Shehata, B., Adarkwah, M. A., Bozkurt, A., Hickey, D. T., Huang, R., & Agyemang, B. (2023). What if the devil is my guardian angel: ChatGPT as a case study of using chatbots in education. Smart Learning Environments, 10, Article 15. https://doi.org/10.1186/s40561-023-00237-x

[13] Miao, F., & Holmes, W. (2023). Guidance for generative AI in education and research. Unesco Publishing. https://unesdoc.unesco.org/ark:/48223/pf0000386693

[14] Yang, Y., Zhang, Y., Wu, M., Zhang, K., Zhang, Y., Yu, H., Hu, Y., & Wang, B. (2025). TwinMarket: A scalable behavioral and social simulation for financial markets [Preprint]. arXiv. https://arxiv.org/abs/2502.01506

[15] Yinxiao Chu (Oct. 1986 - ), male, Han ethnic, Pudong, Shanghai, Assistant Professor, Ph.D., School of Economics, Huazhong University of Science and Technology. Research Interests: market microstructure, financial markets, behavioral finance.  

[16] Xiaoyu Huang (Feb. 1992 - ), female, Han ethnic, Nanchang, Jiangxi, Associate Professor, Ph.D., School of Economics, Wuhan University of Technology. Research interests: asset pricing, macro finance, behavioral finance.