{"paper_id":"2306.11025","title":"Temporal Data Meets LLM -- Explainable Financial Time Series Forecasting","authors":["Xinli Yu","Zheng Chen","Yuan Ling","Shujing Dong","Zongyi Liu","Yanbin Lu"],"paper_date":"2023-06-19","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2306.11025v1","page_url":"https://thequant.space/flowcharts/temporal-data-meets-llm----explainable-financial-time-series/","scores":{"empirical_rigor":7,"math_complexity":4,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes using large language models with structured prompts and fine‑tuning to generate explainable forecasts for NASDAQ‑100 stock returns.","quote":"how LLMs, specifically GPT-4 and Open LLaMA, can be utilized to generate forecasts and provide human‑readable explanations for their predictions.","summary":"This paper presents a novel study on harnessing Large Language Models’ (LLMs) outstanding knowledge and reasoning abilities for explainable financial time series forecasting. The application of machine learning models to financial time series comes with several challenges, including the difficulty…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Price series are transformed into percentage‑change bins for forecasting.","type":"data","stated":"implied","quote":"We first normalize the numeric price time series as a percentage‑change time series, and then categorize the percentage change into bins."},{"assumption":"Weekly and monthly returns are defined as the percentage change from the beginning to the end of the week/month.","type":"data","stated":"implied","quote":"forecasting weekly/monthly stock returns (defined as the percentage change in stock price from the beginning to the end of the week/month)"},{"assumption":"Zero‑shot/few‑shot inference with GPT‑4 can be performed without model fine‑tuning.","type":"model","stated":"explicit","quote":"We demonstrate our structured design of prompts for LLMs and apply the state‑of‑the‑art GPT‑4 model"},{"assumption":"Open LLaMA must be fine‑tuned because it cannot handle instruction‑based prompts in zero‑shot mode.","type":"model","stated":"explicit","quote":"zero‑shot inference, typically tends to replicate portions of the prompt rather than executing the instructions"}]},"datasets":{"status":"available","items":[{"name":"NASDAQ‑100 daily stock price data","provider":"Yahoo Finance","kind":"market-prices","access":"public","period":null,"universe":"NASDAQ‑100 constituents","quote":"price data from Yahoo Finance using yfinance package.","known_id":"yahoo-finance"},{"name":"Company metadata","provider":null,"kind":"fundamentals","access":"public","period":null,"universe":"NASDAQ‑100 constituents","quote":"supplemented by metadata about the stock company"},{"name":"Historical economic/financial news","provider":null,"kind":"text-news","access":"public","period":null,"universe":"NASDAQ‑100 constituents","quote":"concerning both the specific stock and the broader financial/economic landscape"}],"known":[{"id":"alpha-vantage","name":"Alpha Vantage","provider":"Alpha Vantage","kind":"market-prices","access":"public","quote":"(3) The stock sector information, and the stock historical earn-ing obtained from Alpha Vantage (https://www.alphavantage.co/documentation/).","mentions":2,"used_in":1},{"id":"yahoo-finance","name":"Yahoo Finance","provider":"Yahoo","kind":"market-prices","access":"public","quote":"We download daily NASDAQ-100 stock1 price data from Yahoo Finance using yﬁnance package (pypi.org/project/yﬁnance/).","mentions":1,"used_in":1}],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"LLM‑based forecasting outperforms classic ARMA‑GARCH and gradient‑boosting tree baselines.","kind":"performance","quote":"We demonstrate our approach outperforms a few baselines, including the widely applied classic ARMA‑GARCH model and a gradient‑boosting tree model."},{"finding":"Fine‑tuned Open LLaMA achieves reasonable performance but is inferior to GPT‑4.","kind":"performance","quote":"a publicly available LLM such as Open‑LLaMA, after fine‑tuning, can comprehend the instruction to generate explainable forecasts and achieve reasonable performance, albeit relatively inferior in comparison to GPT‑4."},{"finding":"Incorporating Chain‑of‑Thought reasoning significantly improves LLM forecasting performance.","kind":"performance","quote":"Notably, the integration of a step‑by‑step reasoning process, inspired by the Chain of Thought (COT) approach, significantly enhanced the performance of our LLM‑based models."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[],"contradicts":[],"qualifies":[{"paper":{"paper_id":"2603.19944","title":"Large Language Models and Stock Investing: Is the Human Factor Required?","paper_date":"2026-03-20","authors":["Ricardo Crisostomo","Diana Mykhalyuk"],"empirical_rigor":6,"page_url":"https://thequant.space/flowcharts/large-language-models-and-stock-investing-is-the-human-factor-required/"},"question":"Does chain‑of‑thought reasoning improve LLM financial forecasting performance?","this_paper_finding":"Incorporating Chain‑of‑Thought reasoning significantly improves LLM forecasting performance.","other_paper_finding":"Despite improvements, reasoning failures remain pervasive, leading to unstable risk‑adjusted performance and occasional significant mistakes.","other_paper_quote":"Recurring reasoning failures—spanning data retrieval, financial interpretation, computational processing, and meta‑reasoning—persist throughout our live evaluation period.","note":"A reports significant improvement, while B finds pervasive reasoning failures leading to unstable risk‑adjusted results.","confidence":0.557}],"compared_with":22,"citations":{"cited_by_in_corpus":22,"by_independent_authors":20,"how_cited":{"used_as_baseline":0,"uses_method_or_data":1,"extends":0,"discusses_findings":1,"mentions":10,"unclassified":10},"independent_papers":[{"paper":{"paper_id":"2507.07296","title":"Time Series Foundation Models for Multivariate Financial Time Series Forecasting","paper_date":"2025-07-09","authors":["Ben A. Marconi"],"empirical_rigor":7.5,"page_url":"https://thequant.space/flowcharts/time-series-foundation-models-for-multivariate-financial-time-series-forecasting/"},"how":"discusses_findings","quote":"... Another study also showed that such off-the-shelf inference can yield underwhelming results, sometimes falling short of even simple statistical baselines [CITED]."},{"paper":{"paper_id":"2402.03659","title":"Learning to Generate Explainable Stock Predictions using Self-Reflective Large Language Models","paper_date":"2024-02-06","authors":["Kelvin J. L. Koa","Yunshan Ma","Ritchie Ng","Tat-Seng Chua"],"empirical_rigor":8,"page_url":"https://thequant.space/flowcharts/learning-to-generate-explainable-stock-predictions-using-sel/"},"how":"uses_method_or_data","quote":null}],"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":"2026-11-14T18:25:35Z"},"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/"}