{"paper_id":"2409.08282","title":"LSR-IGRU: Stock Trend Prediction Based on Long Short-Term Relationships and Improved GRU","authors":["Peng Zhu","Yuante Li","Yifan Hu","Qinyuan Liu","Dawei Cheng","Yuqi Liang"],"paper_date":"2024-08-26","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2409.08282v3","page_url":"https://thequant.space/flowcharts/lsr-igru-stock-trend-prediction-based-on-long-short-term-re/","scores":{"empirical_rigor":6,"math_complexity":7,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes LSR-IGRU, a stock price trend prediction model that integrates long‑term and short‑term stock relationships with improved GRU inputs.","quote":"we propose a stock price trend prediction model named LSR-IGRU in this paper, which is based on long short-term stock relationships and an improved GRU input.","summary":"Stock price prediction is a challenging problem in the field of finance and receives widespread attention. In recent years, with the rapid development of technologies such as deep learning and graph neural networks, more research methods have begun to focus on exploring the interrelationships…","strategy":"LSR-IGRU"},"assumptions":{"status":"none_found","items":[]},"datasets":{"status":"available","items":[{"name":"CSI 300","provider":null,"kind":"market-prices","access":"unknown","period":"2018-2023","universe":null,"quote":"We conduct experiments using data from both the Chinese and American stock markets. These datasets include CSI 300, CSI 500, S&P 500, and NASDAQ 100."},{"name":"CSI 500","provider":null,"kind":"market-prices","access":"unknown","period":"2018-2023","universe":null,"quote":"We conduct experiments using data from both the Chinese and American stock markets. These datasets include CSI 300, CSI 500, S&P 500, and NASDAQ 100."},{"name":"S&P 500","provider":null,"kind":"market-prices","access":"unknown","period":"2018-2023","universe":null,"quote":"We conduct experiments using data from both the Chinese and American stock markets. These datasets include CSI 300, CSI 500, S&P 500, and NASDAQ 100."},{"name":"NASDAQ 100","provider":null,"kind":"market-prices","access":"unknown","period":"2018-2023","universe":null,"quote":"We conduct experiments using data from both the Chinese and American stock markets. These datasets include CSI 300, CSI 500, S&P 500, and NASDAQ 100."}],"known":[],"sample":{"start":"2023","end":"2023","market":"CSI 300 stocks","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":"unclear","evidence_card_quoted":false},"meaning":"a trading or execution result is reported without costs"},"results":{"status":"available","evaluation_type":"trading-pnl","findings":[{"finding":"LSR-IGRU outperforms current state‑of‑the‑art baseline models on all four datasets.","kind":"performance","quote":"we validate the superiority of the proposed LSR-IGRU model over the current state-of-the-art baseline models."},{"finding":"When applied to a financial company’s algorithmic trading system, LSR-IGRU achieves significantly higher cumulative portfolio returns than other baselines.","kind":"performance","quote":"achieving significantly higher cumulative portfolio returns compared to other baseline methods."},{"finding":"Back‑test results show the model’s returns exceed market indices and maintain a low excess‑return drawdown rate of about 5% in the worst case.","kind":"performance","quote":"our model maintains a very low drawdown rate over a long period, reaching only about 5% in the worst case."},{"finding":"Ablation studies indicate that each of the three modules (GRU, long‑term relationship, short‑term relationship) contributes to performance, with the full model achieving the best results.","kind":"predictive-accuracy","quote":"using any single module alone failed to achieve the optimal performance level, while combining any two modules exhibited stronger predictive accuracy."},{"finding":"Parameter sensitivity experiments show ARR and IR improve up to a 15‑day time window and then decline, and optimal performance is achieved with 2 GCN layers and limited GRU layers.","kind":"statistical","quote":"Figures 2(a) and 2(e) show ARR and IR increasing with the time step up to 15 days, then declining."}],"reported_performance":{"strategy":"LSR-IGRU","sharpe":1.026,"sharpe_benchmark":null,"annual_return_pct":0.192,"total_return_pct":null,"max_drawdown_pct":0.168},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":30,"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":1,"unclassified":1},"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/ZP1481616577/Baselines_LSR-IGRU","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":"2026-11-14T18:03:15Z"},"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/"}