{"paper_id":"2601.03948","title":"Trade-R1: Bridging Verifiable Rewards to Stochastic Environments via Process-Level Reasoning Verification","authors":["Rui Sun","Yifan Sun","Sheng Xu","Li Zhao","Jing Li","Daxin Jiang","Cheng Hua","Zuo Bai"],"paper_date":"2026-01-07","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2601.03948v2","page_url":"https://thequant.space/flowcharts/trade-r1-bridging-verifiable-rewards-to-stochastic-environm/","scores":{"empirical_rigor":6,"math_complexity":6,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes Trade-R1, a framework that bridges verifiable rewards to stochastic financial environments via process-level reasoning verification and semantic reward strategies.","quote":"We presented Trade-R1, a framework bridging verifiable rewards to stochastic environments via process-level reasoning verification.","summary":"Reinforcement Learning (RL) has enabled Large Language Models (LLMs) to achieve remarkable reasoning in domains like mathematics and coding, where verifiable rewards provide clear signals. However, extending this paradigm to financial decision is challenged by the market’s stochastic nature…","strategy":"DSR (Dynamic-effect Semantic Reward)"},"assumptions":{"status":"available","items":[{"assumption":"The market reward r is defined as the 10‑day forward excess return relative to the local benchmark.","type":"market","stated":"explicit","quote":"The market reward r is defined as the 10‑day forward excess return relative to the local benchmark (CSI 300)."},{"assumption":"The policy model generates a reasoning chain c and a decision d, and optimization maximizes a semantic‑gated reward G(r, s).","type":"model","stated":"explicit","quote":"... our optimization objective maximizes the semantic‑gated reward: J(θ) = Ex∼D,y∼πθ(·|x)[G(r, s)]"},{"assumption":"Semantic similarity score s ∈[0,1] can be computed and used as a gating factor.","type":"model","stated":"explicit","quote":"where G is our proposed asymmetric gating function and s ∈[0, 1] is the semantic similarity score."},{"assumption":"Training uses a fixed‑size Qwen3‑8B model and textual news inputs only.","type":"model","stated":"explicit","quote":"We employ Qwen3‑8B‑Instruct as the policy model for both markets, optimized via full‑parameter training."},{"assumption":"The backtesting window is limited to the LLM’s knowledge cutoff to avoid data leakage.","type":"data","stated":"explicit","quote":"Our backtesting window is constrained by the LLM’s knowledge cutoff to prevent data leakage."},{"assumption":"Training is stopped at a pre‑defined step due to computational budget, not at full convergence.","type":"execution","stated":"explicit","quote":"Due to computational budget, we terminated training at a pre‑defined step rather than waiting for complete convergence."}]},"datasets":{"status":"available","items":[{"name":"A‑share Market (CN)","provider":null,"kind":"text-news","access":"unknown","period":"July 2024 to October 2025","universe":"All A‑share stocks","quote":"A‑share Market (CN): Constructed from daily Chinese financial news summaries. The target universe covers all A‑share stocks."},{"name":"US Stock Market (US)","provider":null,"kind":"text-news","access":"unknown","period":"July 2024 to October 2025","universe":"All stocks listed on major exchanges (NYSE, NASDAQ)","quote":"US Stock Market (US): Constructed from daily English financial news summaries. The target universe covers all stocks listed on major exchanges (e.g., NYSE, NASDAQ), explicitly excluding OTC securities to ensure liquidity."}],"known":[],"sample":{"start":"2025-07","end":"2025-10","market":"A-Share (CSI-300)","asset_class":"equities","frequency":"daily"},"data_resolution":"daily"},"transaction_costs":{"status":"available","modeled":"no","headline_net_of_costs":false,"detail":null,"quote":null,"source":"front_matter","rule":"R1","candidates":{"front_matter":"no","evidence_card":"no","evidence_card_quoted":false},"meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"Dynamic‑effect Semantic Reward (DSR) achieves the highest semantic similarity score (0.9744) and the lowest hallucination rate (0.0012) on the A‑share test set, while also delivering strong returns.","kind":"performance","quote":"... DSR achieves: (1) the highest Semantic Similarity Score (0.9744), indicating superior reasoning faithfulness; and (2) the lowest Hallucination Rate (0.0012), demonstrating factual grounding."},{"finding":"On the out‑of‑distribution US market, DSR attains higher cumulative return (15.34%) and a better Sharpe ratio (1.951) than both Market‑Only and FSR, along with the highest similarity score (0.7768).","kind":"performance","quote":"On the out‑of‑distribution US Market, DSR demonstrates superior cross‑market generalization: it achieves higher returns (15.34%) and a better Sharpe ratio (1.951) than both Market‑Only and FSR, alongside the highest similarity score (0.7768)."},{"finding":"Fixed‑effect Semantic Reward (FSR) yields a cumulative return of 39.38% on the A‑share test set, slightly lower than Market‑Only but with stable alignment signals.","kind":"performance","quote":"... cumulative return (37.76%) is slightly lower than FSR, DSR achieves"},{"finding":"Unconstrained market‑only RL achieves competitive financial returns (37.62% Cum. Ret.) on A‑share but suffers from low reasoning alignment (semantic similarity 0.4369, hallucination 0.2254).","kind":"performance","quote":"achieves competitive financial returns on the A‑Share market (37.62% Cum. Ret.) but exhibits"},{"finding":"The proposed Retrieval‑Augmented Verification with a Triangular Consistency metric enables accurate process supervision over long financial documents.","kind":"theoretical","quote":"We introduced a Retrieval‑Augmented Verification protocol with a Triangular Consistency metric, assessing Factuality, Deduction, and Consistency through pairwise alignment between evidence, reasoning chains, and decisions, enabling accurate process supervision over long financial documents."}],"reported_performance":{"strategy":"DSR (Dynamic-effect Semantic Reward)","sharpe":3.036,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":37.76,"max_drawdown_pct":12.04},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":43,"citations":{"cited_by_in_corpus":2,"by_independent_authors":1,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":2},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"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":"2026-11-14T17:17:05Z"},"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/"}