aiqbench analysis
This research critically assesses the limitations of using Large Language Models as personalized user simulators, specifically within financial trading contexts. The study finds that current LLMs fail to reliably reproduce individual decision logic over time, performing no better than simple persistence baselines and struggling with asset selection fidelity. These results highlight a significant gap between coarse behavioral agreement and true individual-level fidelity, suggesting that future evaluation frameworks must separately address explanation alignment and evidence dependence in personalized simulations.
Beyond Behavioral Accuracy: Evaluating Individual-Level Fidelity in LLM Investor Simulation
arXiv:2610.26524v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as user simulators, but can they faithfully reproduce an individual's behavior and decision logic over time? We investigate this question in a longitudinal paper-trading study with 80 participants. Using rolling next-day predictions, we evaluate trading activity, buy-sell structure, asset selection, and portfolio outcomes, complemented by controlled input ablations and participant-rated explanation alignment. No evaluated LLM reliably outperforms a simple recent-activity persistence baseline for predicting trade occurrence, while buy-sell structure and asset selection remain difficult to reproduce. Similar activity-level performance can also yield substantially different portfolio trajectories. Ablations show that activity predictions depend primarily on recent trades, whereas asset choices are sensitive to multiple evidence sources. Participants rate explanations as more aligned with their recurring strategies than with their decision-specific rationales. These findings reveal a gap between coarse behavioral agreement and individual-level fidelity, motivating separate evaluation of behavioral accuracy, explanation alignment, and evidence dependence in personalized user simulation.
Abstract:Large language models (LLMs) are increasingly used as user simulators, but can they faithfully reproduce an individual's behavior and decision logic over time? We investigate this question in a longitudinal paper-trading study with 80 participants. Using rolling next-day predictions, we evaluate trading activity, buy-sell structure, asset selection, and portfolio outcomes, complemented by controlled input ablations and participant-rated explanation alignment. No evaluated LLM reliably outperforms a simple recent-activity persistence baseline for predicting trade occurrence, while buy-sell structure and asset selection remain difficult to reproduce. Similar activity-level performance can also yield substantially different portfolio trajectories. Ablations show that activity predictions depend primarily on recent trades, whereas asset choices are sensitive to multiple evidence sources. Participants rate explanations as more aligned with their recurring strategies than with their decision-specific rationales. These findings reveal a gap between coarse behavioral agreement and individual-level fidelity, motivating separate evaluation of behavioral accuracy, explanation alignment, and evidence dependence in personalized user simulation.
Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Information Retrieval (cs.IR)
arXiv:2610.26524 [cs.AI]
(or arXiv:2610.26524v1 [cs.AI] for this version)
https://doi.org/10.48550/arXiv.2610.26524 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 252 words