LLM agents are promising tools for empirical discovery, but their flexibility can also turn discovery into uncontrolled search. We study how to use agents under a reproducible protocol through cryptocurrency factor discovery. Our framework casts the task as sequential hypothesis search: an agent reads an append-only experiment trace, proposes falsifiable factor hypotheses, and maps them to executable recipes, while a deterministic engine enforces fixed data splits, selection gates, transaction costs, and portfolio tests. Candidate actions are restricted to a point-in-time factor DSL, making both successful and failed hypotheses auditable. A ridge-combined portfolio trained only on 2020--2022 data achieves a 44.55% annualized return and Sharpe ratio of 1.55 in the 2024--2026 pure out-of-sample period after a 5 basis point one-way trading cost.
Cite
@article{arxiv.2604.26747,
title = {From Hypotheses to Factors: Constrained LLM Agents in Cryptocurrency Markets},
author = {Yikuan Huang and Zheqi Fan and Kaiqi Hu and Yifan Ye},
journal= {arXiv preprint arXiv:2604.26747},
year = {2026}
}