{"paper_id":"2508.00554","title":"ContestTrade: A Multi-Agent Trading System Based on Internal Contest Mechanism","authors":["Li Zhao","Rui Sun","Zuoyou Jiang","Bo Yang","Yuxiao Bai","Mengting Chen","Xinyang Wang","Jing Li","Zuo Bai"],"paper_date":"2025-08-01","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2508.00554v3","page_url":"https://thequant.space/flowcharts/contesttrade-a-multi-agent-trading-system-based-on-internal/","scores":{"empirical_rigor":6,"math_complexity":3,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes ContestTrade, a novel multi‑agent trading system with internal competitive mechanisms and a Deep Research methodology.","quote":"To address these limitations, we propose ContestTrade (A Multi-Agent Trading System Based on Internal Contest Mechanism).","summary":"In financial trading, large language model (LLM)-based agents demonstrate significant potential. However, the high sensitivity to market noise undermines the performance of LLM-based trading systems. To address this limitation, we propose a novel multi-agent system featuring an internal competitive…","strategy":"ContestTrade"},"assumptions":{"status":"available","items":[{"assumption":"Factor performance exhibits short‑term momentum.","type":"market","stated":"explicit","quote":"We hypothesize that factor performance exhibits short‑term momentum."},{"assumption":"Optimal short‑term windows are m = 5 days and n = 3 days for factor prediction.","type":"statistical","stated":"explicit","quote":"with optimal values found at m = 5, n = 3"},{"assumption":"Testing period (January‑June 2025) is after the LLM knowledge cutoff, eliminating data leakage.","type":"data","stated":"explicit","quote":"The testing period (January-June 2025) is chosen to be entirely after the knowledge cutoff of our LLMs, eliminating potential leakage from their pre‑training data."},{"assumption":"Agents can be evaluated using a hybrid score combining realized Sharpe Ratio over a trailing m‑day window and qualitative judgments from an LLM Judger Panel.","type":"model","stated":"explicit","quote":"Realized Performance: Standard quantitative metrics calculated over a trailing m‑day window (e.g., realized Sharpe Ratio)."},{"assumption":"LightGBM is an appropriate model for mapping historical hybrid scores to future utility.","type":"model","stated":"explicit","quote":"we employ LightGBM for this task, reusing the same simple and robust setup."}]},"datasets":{"status":"available","items":[{"name":"real‑world financial dataset (news, corporate financials, market data)","provider":null,"kind":"text-news","access":"unknown","period":"training: July‑December 2024; testing: J","universe":null,"quote":"Our experiments utilize a real‑world financial dataset encompassing news, corporate financials, and market data."}],"known":[],"sample":{"start":"2025-01","end":"2025-06","market":null,"asset_class":"multi-asset","frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"partial","headline_net_of_costs":false,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"partial","evidence_card":"unclear","evidence_card_quoted":false},"meaning":"costs appear only as a robustness check, a flat fee without spread or impact, or outside the headline numbers"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"ContestTrade achieves a cumulative return of 52.80%, Sharpe ratio of 3.12, and maximum drawdown of 12.41%, outperforming all baselines.","kind":"performance","quote":"It achieves a Cumulative Return (CR) of 52.80%, a Sharpe Ratio (SR) of 3.12, and a Maximum Drawdown (MDD) of only 12.41%."},{"finding":"Ablation shows removing the LLM Judge reduces CR to 50.55%, SR to 2.57, and MDD to 13.48%; removing the contest in the Research team drops CR to 32.83%, SR to 1.78, MDD to 16.70%.","kind":"performance","quote":"w/o LLM Judge 50.55 2.57 13.48 ... w/o Contest - Researcher 32.83 1.78 16.70"},{"finding":"Internal contest mechanisms yield high Rank IC and ICIR, indicating strong predictive quality of selected factors and signals.","kind":"predictive-accuracy","quote":"The consistently high scores in these metrics demonstrate that our mechanism effectively identifies high‑quality inputs, which in turn is the primary driver of the final portfolio’s outperformance."}],"reported_performance":{"strategy":"ContestTrade","sharpe":3.12,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":52.8,"max_drawdown_pct":12.41},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":44,"citations":{"cited_by_in_corpus":3,"by_independent_authors":3,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":3,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/FinStep-AI/ContestTrade","note":null},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T16:37:25Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}