{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,18]],"date-time":"2026-10-18T15:39:35Z","timestamp":1792337975731,"version":"4.4.0"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":44,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/www.uspto.gov\/ip-policy\/copyright-policy\/copyright-basics"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>Large language models (LLMs) are increasingly used in algorithmic trading, yet knowledge integration across heterogeneous evidence remains underexplored in emerging and non-English markets. We present a five-agent knowledge-fusion framework for financial decision support in Vietnamese equities. The architecture integrates technical indicators, point-in-time quantitative alpha selection, financial sentiment, and vision-language chart analysis, providing a reproducible basis for combining heterogeneous quantitative, textual, and visual evidence in multi-agent reasoning. A paired shared report protocol isolates the contribution of alpha information while holding upstream evidence fixed, and a four-way ablation separates quantitative-factor and sentiment effects. Across nine equities and 180 paired decisions, the Full System attains 65.0% accuracy versus 51.1% without Alpha. The ablation identifies a larger independent contribution from alpha factors than from sentiment, with stock-dependent interactions. An execution-aware long-only ledger nevertheless produces a negative mean net return, showing that improved directional inference does not ensure economic profitability. The framework therefore supports auditable knowledge integration and controlled component evaluation for financial decision support; claims of immediate trading deployment remain outside the available evidence.<\/jats:p>","DOI":"10.2139\/ssrn.7629484","type":"posted-content","created":{"date-parts":[[2026,10,18]],"date-time":"2026-10-18T14:37:26Z","timestamp":1792334246000},"source":"Crossref","is-referenced-by-count":0,"title":["Enhancing Multi Agent Large Language Models for Stock Trading through Quantitative Alpha Integration"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4675-2123","authenticated-orcid":true,"given":"Viet  Hong","family":"Tran","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-7266-2437","authenticated-orcid":true,"given":"Van-Tang","family":"Nguyen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hoang-Dung","family":"Phan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"ref1","journal-title":"ONLY return the 6 fields below. Do NOT add any other fields"},{"key":"ref2","journal-title":"NO explanations"},{"key":"ref3","journal-title":"NO fallback scenarios. Provide exactly ONE definitive assessment"},{"key":"ref4","journal-title":"MANDATORY FORMAT: **Pattern:** <Name> **Confidence:** High | Medium | Low **Forecast Bias:** Bullish | Bearish | Neutral **Evidence:** <Brief description> **Key Candle:** <List key candle> **"},{"key":"ref5","journal-title":"Provide a short comment (1-2 sentences) on current momentum"},{"key":"ref6","journal-title":"Synthesize: Are flow and sentiment in agreement or contradiction?"},{"key":"ref7","journal-title":"Identify hidden risks\/opportunities from news not captured by technical alphas. Do NOT over-analyze. Deduce from the numbers to conclude the view for {horizon_label}"},{"key":"ref8","author":"F Xiong","year":"2025","journal-title":"Quantagent: Price-driven multi-agent llms for high-frequency trading"},{"key":"ref9","author":"A I Langchain","year":"2024"},{"key":"ref10","author":"H Yang","year":"2023","journal-title":"Fingpt: Open-source financial large language models"},{"key":"ref11","author":"S Wu","year":"2023","journal-title":"Bloomberggpt: A large language model for finance"},{"key":"ref12","author":"T Research","year":"2024","journal-title":"Tradingagents: Multi-agents llm financial trading framework"},{"key":"ref13","first-page":"7","article-title":"Forecasting stock prices from the limit order book using convolutional neural networks","author":"A Tsantekidis","year":"2017","journal-title":"IEEE 19th Conference on Business Informatics"},{"issue":"3","key":"ref14","doi-asserted-by":"crossref","first-page":"653","DOI":"10.1109\/TNNLS.2016.2522401","article-title":"Deep direct reinforcement learning for financial signal representation and trading","volume":"28","author":"Y Deng","year":"2016","journal-title":"IEEE Transactions on Neural Networks and Learning Systems"},{"issue":"4","key":"ref15","doi-asserted-by":"crossref","first-page":"1705","DOI":"10.1111\/0022-1082.00265","article-title":"Foundations of technical analysis: Computational algorithms, statistical inference, and empirical implementation","volume":"55","author":"A W Lo","year":"2000","journal-title":"The Journal of Finance"},{"key":"ref16","author":"Y Han","year":"2023","journal-title":"Chartllama: A multimodal llm for chart understanding and generation"},{"key":"ref17","doi-asserted-by":"crossref","first-page":"13447","DOI":"10.18653\/v1\/2026.acl-long.615","article-title":"Finchart-bench: Benchmarking financial chart comprehension in vision-language models","volume":"1","author":"D Shu","year":"2026","journal-title":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics"},{"key":"ref18","author":"Z Wang","year":"2026","journal-title":"Martingale doppelg\u00e4nger-eval: An identification framework for auditing candlestick understanding in vision-language models"},{"issue":"1","key":"ref19","doi-asserted-by":"crossref","first-page":"65","DOI":"10.1111\/j.1540-6261.1993.tb04702.x","article-title":"Returns to buying winners and selling losers: Implications for stock market efficiency","volume":"48","author":"N Jegadeesh","year":"1993","journal-title":"The Journal of Finance"},{"issue":"2","key":"ref20","first-page":"427","article-title":"The cross-section of expected stock returns","volume":"47","author":"E F Fama","year":"1992","journal-title":"The Journal of Finance"},{"key":"ref21","article-title":"Single-stock alpha factors based on price-volume features","volume":"168","author":"J Kang","year":"2021","journal-title":"Expert Systems with Applications"},{"issue":"1","key":"ref22","doi-asserted-by":"crossref","first-page":"5","DOI":"10.1093\/rfs\/hhv059","article-title":"and the cross-section of expected returns","volume":"29","author":"C R Harvey","year":"2016","journal-title":"The Review of Financial Studies"},{"issue":"1","key":"ref23","doi-asserted-by":"crossref","first-page":"15","DOI":"10.3905\/jpm.2017.44.1.015","article-title":"A century of evidence on trend-following investing","volume":"44","author":"B Hurst","year":"2017","journal-title":"The Journal of Portfolio Management"},{"key":"ref24","doi-asserted-by":"crossref","DOI":"10.1186\/1471-2105-7-91","article-title":"Bias in error estimation when using cross-validation for model selection","volume":"7","author":"S Varma","year":"2006","journal-title":"BMC Bioinformatics"},{"issue":"5","key":"ref25","doi-asserted-by":"crossref","first-page":"94","DOI":"10.3905\/jpm.2014.40.5.094","article-title":"The deflated sharpe ratio: Correcting for selection bias, backtest overfitting, and non-normality","volume":"40","author":"D H Bailey","year":"2014","journal-title":"The Journal of Portfolio Management"},{"issue":"3","key":"ref26","doi-asserted-by":"crossref","first-page":"1139","DOI":"10.1111\/j.1540-6261.2007.01232.x","article-title":"Giving content to investor sentiment: The role of media in the stock market","volume":"62","author":"P C Tetlock","year":"2007","journal-title":"The Journal of Finance"},{"key":"ref27","author":"D Araci","year":"2019","journal-title":"Finbert: Financial sentiment analysis with pre-trained language models"},{"issue":"2","key":"ref28","article-title":"Market efficiency in vietnam's stock market: Evidence from a weak-form test","volume":"12","author":"X V Vo","year":"2019","journal-title":"Journal of Risk and Financial Management"},{"issue":"3","key":"ref29","first-page":"45","article-title":"Herding behavior in emerging equity markets: Evidence from vietnam","volume":"8","author":"T H V Pham","year":"2021","journal-title":"The Journal of Asian Finance, Economics and Business"},{"issue":"4","key":"ref30","first-page":"12","article-title":"Technical trading rules in the vietnamese stock market: Empirical evidence under price limits","volume":"17","author":"H T Truong","year":"2020","journal-title":"Investment Management and Financial Innovations"},{"key":"ref31","author":"S Fatemi","year":"2024","journal-title":"Finvision: A multi-agent framework for stock market prediction"},{"key":"ref32","author":"Y Xia","year":"2026","journal-title":"Agentic trading: When llm agents meet financial markets"},{"key":"ref33","author":"J Zhang","year":"2026","journal-title":"Prediction arena: Benchmarking ai models on real-world prediction markets"},{"key":"ref34","author":"A Arora","year":"2026","journal-title":"Predictionmarketbench: A swe-bench-style framework for backtesting trading agents on prediction markets"},{"key":"ref35","author":"Z Zeng","year":"2025","journal-title":"Futurex: An advanced live benchmark for llm agents in future prediction"},{"key":"ref36","author":"R C Grinold","year":"1999","journal-title":"Active Portfolio Management"},{"issue":"1","key":"ref37","doi-asserted-by":"crossref","first-page":"119","DOI":"10.1086\/294846","article-title":"Mutual fund performance","volume":"39","author":"W F Sharpe","year":"1966","journal-title":"Journal of Business"},{"key":"ref38","author":"J Wei","year":"2022","journal-title":"Chain-of-thought prompting elicits reasoning in large language models"},{"key":"ref39","author":"R Pardo","year":"2008","journal-title":"The Evaluation and Optimization of Trading Strategies"},{"key":"ref40","author":"G Appel","year":"2005","journal-title":"Technical Analysis: Power Tools for Active Investors"},{"key":"ref41","author":"J W Wilder","year":"1978","journal-title":"New Concepts in Technical Trading Systems"},{"key":"ref42","author":"G Lane","year":"1984","journal-title":"Lane's stochastic oscillator"},{"key":"ref43","author":"L Williams","year":"1973","journal-title":"Trading stocks and commodities with the insiders"},{"key":"ref44","author":"J J Murphy","year":"1999","journal-title":"Technical Analysis of the Financial Markets"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,18]],"date-time":"2026-10-18T14:40:19Z","timestamp":1792334419000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7629484"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":44,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7629484","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}