FinMem: A Performance-Enhanced LLM Trading Agent With Layered Memory and Character Design
The DOI matches its Crossref record; 63 of the 63 references it cites were found in a registry or an index.
Abstract
We introduceFinMem, a novel Large Language Models (LLM)-based agent framework for financial trading, designed to address the need for automated systems that can transform real-time data into executable decisions.FinMemcomprises three core modules: Profile for customizing agent characteristics, Memory for hierarchical financial data assimilation, and Decision-making for converting insights into investment choices. The Memory module, which mimics human traders' cognitive structure, offers interpretability and real-time tuning while handling the critical timing of various information types. It employs a layered approach to process and prioritize data based on its timeliness and relevance, ensuring that the most recent and impactful information is given appropriate weight in decision-making.FinMem's adjustable cognitive span allows retention of critical information beyond human limits, enabling it to balance historical patterns with current market dynamics. This framework facilitates self-evolution of professional knowledge, agile reactions to investment cues, and continuous refinement of trading decisions in financial environments. When compared against advanced algorithmic agents using a large-scale real-world financial dataset,FinMemdemonstrates superior performance across classic metrics like Cumulative Return and Sharpe ratio. Further tuning of the agent's perceptual span and character setting enhances its trading performance, positioningFinMemas a cutting-edge solution for automated trading.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Citations
Cites
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
The magical number seven, plus or minus two: Some limits on our capacity for processing information.
Miller, G. A. (1956) · Psychological Review · doi:10.1037/h0043158 · Cited by 17,812 · cited 1×
“Cognitive research suggests that human working memory is limited to recalling five to nine events at once (Miller 1956).”
BackgroundLevels of processing: A framework for memory research
Craik, F. I. M., & Lockhart, R. S. (1972) · Journal of Verbal Learning and Verbal Behavior · doi:10.1016/s0022-5371(72)80001-x · Cited by 9,566
Asynchronous Methods for Deep Reinforcement Learning
Mnih, V., Badia, A. P., Mirza, M., Graves, A., Lillicrap, T. P., Harley, T., Silver, D., & Kavukcuoglu, K. (2016) · arXiv · doi:10.48550/arxiv.1602.01783 · Cited by 7,132
Billion-scale similarity search with GPUs
Johnson, J., Douze, M., & Jegou, H. (2021) · IEEE Transactions on Big Data · doi:10.1109/tbdata.2019.2921572 · Cited by 4,228
Billion-scale similarity search with GPUs
Johnson, J., Douze, M., & Jégou, H. (2017) · arXiv · doi:10.48550/arxiv.1702.08734 · Cited by 4,140
MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework
Hong, S., Zhuge, M., Chen, J., Zheng, X., Cheng, Y., Zhang, C., Wang, J., Wang, Z., Yau, S. K. S., Lin, Z., et al. (2023) · arXiv · doi:10.48550/arxiv.2308.00352 · Cited by 2,825
Sharpe, W. F. (1994) · The Journal of Portfolio Management · doi:10.3905/jpm.1994.409501 · Cited by 2,534
Generative Agents: Interactive Simulacra of Human Behavior
Park, J. S., O'Brien, J., Cai, C. J., Morris, M. R., Liang, P., & Bernstein, M. S. (2023) · Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology · doi:10.1145/3586183.3606763 · Cited by 2,187
Generative Agents: Interactive Simulacra of Human Behavior
Park, J. S., O'Brien, J. C., Cai, C. J., Morris, M. R., Liang, P., & Bernstein, M. S. (2023) · arXiv · doi:10.48550/arxiv.2304.03442 · Cited by 2,173
Research works printed but not linked: 6
- Chen, 2012, Forecasting the Taiwan stock market with a novel momentum-based fuzzy time-series, Rev. Econ. Finance, 2, 38
- 10.1145/3586183.3606763 · DOI · Paper page
- 10.1145/3490354.3494366 · DOI
- Chung, 2013, Playing atari with deep reinforcement learning, Comput. Ence, 21
- Hull, 2007, Risk Management and Financial Institutions
- Raffin, 2021, Stable-baselines3: Reliable reinforcement learning implementations, 22, 1
Cited by
Capable language models can outgrow the benefits of collaboration
Kim, Y., Gu, K., Park, C., Park, C., Schmidgall, S., Heydari, A. A., Yan, Y., Zhang, Z., Zhuang, Y., Liu, Y., et al. (2026) · Nature Machine Intelligence · doi:10.1038/s42256-026-01268-y · Cited by 27
“Independent architecture’s universal underperformance (mean success 0.370 versus 0.466 SAS) stems from the absence of interagent 9 and 10, not by this table alone.”
InfluentialMem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
Chhikara, P., Khant, D., Aryan, S., Singh, T., & Yadav, D. (2025) · arXiv · doi:10.48550/arxiv.2504.19413 · Cited by 10
TradingAgents: Multi-Agents LLM Financial Trading Framework
Xiao, Y., Sun, E., Luo, D., & Wang, W. (2024) · arXiv · doi:10.48550/arxiv.2412.20138 · Cited by 6
Empowered stock market forecasting using Large Language Model on social media content
Sperlí, G., & Sichinolfi, M. A. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.112727 · Cited by 13
Toward Sustainable Agentic AI Systems: A Survey of Architectures and Methodologies
Lee, Y., & Park, E. (2026) · Sustainable Development · doi:10.1002/sd.70942 · Cited by 13
Metadata
- Authors
- Yu, Y., Li, H., Chen, Z., Jiang, Y., Li, Y., Suchow, J. W., Zhang, D., & Khashanah, K.
- Affiliations
- Stevens Institute of Technology (US)
- ORCID
- Venue
- IEEE Transactions on Big Data
- Published
- 2025-12
- Volume
- 11
- Issue
- 6
- Pages
- 3443-3459
- ISSN
- 2332-7790 (ISSN-L)
- 2372-2096
- OpenAlex
- W4412939561
- Cited by
- 23
- References
- 50
- Citations by year
- 2026: 20 · 2025: 3
- FWCI
- 17.60
- Type
- journal article
- Language
- en
- Open access
- closed
- Licence
- Locations
- IEEE Transactions on Big Data (publishedVersion)
- Topics
- Social Sciences › Decision Sciences › Management Science and Operations Research › Stock Market Forecasting Methods
- Social Sciences › Economics, Econometrics and Finance › Finance › Financial Markets and Investment Strategies
- Social Sciences › Business, Management and Accounting › Management Information Systems › FinTech, Crowdfunding, Digital Finance
- Keywords
- algorithmic trading, financial trading, real-time data processing, investment decision-making, interpretability, Sharpe ratio, cumulative returns
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.