From Hypotheses to Factors: Constrained LLM Agents in Cryptocurrency Markets
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909000849686528 |
|---|---|
| author | Huang, Yikuan Fan, Zheqi Hu, Kaiqi Ye, Yifan |
| author_facet | Huang, Yikuan Fan, Zheqi Hu, Kaiqi Ye, Yifan |
| contents | 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. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_26747 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | From Hypotheses to Factors: Constrained LLM Agents in Cryptocurrency Markets Huang, Yikuan Fan, Zheqi Hu, Kaiqi Ye, Yifan Portfolio Management General Finance Trading and Market Microstructure 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. |
| title | From Hypotheses to Factors: Constrained LLM Agents in Cryptocurrency Markets |
| topic | Portfolio Management General Finance Trading and Market Microstructure |
| url | https://arxiv.org/abs/2604.26747 |