{"paper_id":"2409.11408","title":"Optimizing Performance: How Compact Models Match or Exceed GPT's Classification Capabilities through Fine-Tuning","authors":["Baptiste Lefort","Eric Benhamou","Jean-Jacques Ohana","David Saltiel","Beatrice Guez"],"paper_date":"2024-08-22","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2409.11408v1","page_url":"https://thequant.space/flowcharts/optimizing-performance-how-compact-models-match-or-exceed-g/","scores":{"empirical_rigor":7,"math_complexity":4,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes a market‑based dataset and fine‑tunes compact models (FinBERT, FinDRoBERTa) to outperform GPT‑3.5/4 on financial sentiment classification.","quote":"We fine‑tune Fin‑BERT, FinDRoBERTa and GPT‑3.5 on the market‑based dataset.","summary":"In this paper, we demonstrate that non-generative, small-sized models such as FinBERT and FinDRoBERTa, when fine-tuned, can outperform GPT-3.5 and GPT-4 models in zero-shot learning settings in sentiment analysis for financial news. These fine-tuned models show comparable results to GPT-3.5 when it…","strategy":"sentiment signal equity strategy"},"assumptions":{"status":"available","items":[{"assumption":"The market‑based labels reflect the true market reaction, not human annotator bias.","type":"data","stated":"explicit","quote":"The label must reflect the market’s true behavior, not just the human annotator’s assessment."},{"assumption":"All individual classifiers are better than random guessing.","type":"statistical","stated":"explicit","quote":"The key assumption beyond independence is that all the individual classifiers, Cn, are a better than random guessing."},{"assumption":"Models are independent estimators for Condorcet’s theorem to hold.","type":"statistical","stated":"implied","quote":"If several independent estimators, that are better than random decision, are combined into a majority vote classifier"},{"assumption":"Fine‑tuning uses identical parameters and the same dataset for all models.","type":"model","stated":"explicit","quote":"All models underwent fine‑tuning with identical parameters and the same dataset."}]},"datasets":{"status":"available","items":[{"name":"market‑based dataset (financial_headlines_market_based)","provider":"Bloomberg (via HuggingFace)","kind":"text-news","access":"public","period":"2010‑2024","universe":null,"quote":"we developed a new and large dataset derived from all major Bloomberg news from 2010 to 2024 that assigns market reaction to news headlines while eliminating human bias","known_id":"bloomberg"}],"known":[{"id":"bloomberg","name":"Bloomberg","provider":"Bloomberg L.P.","kind":"market-prices","access":"commercial","quote":"These fine-tuned models show comparable results to GPT-3.5 when it is fine-tuned on the task of determining market sentiment from daily financial news summaries sourced from Bloomberg.","mentions":7,"used_in":3}],"sample":{"start":null,"end":null,"market":"main equity markets","asset_class":"equities","frequency":"daily"},"data_resolution":"daily"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"evidence_card","rule":"R2","candidates":{"front_matter":"unclear","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":"Compact, non‑generative models fine‑tuned on the market‑based dataset surpass the zero‑shot capabilities of GPT‑3.5 and GPT‑4 in financial sentiment analysis.","kind":"performance","quote":"we illustrate that with proper fine‑tuning, compact, non‑generative models such as FinBERT and FinDRoBERTa are able to surpass the zero‑shot learning capabilities of larger models like GPT‑3.5 and GPT‑4"},{"finding":"Bagging (majority‑vote) does not improve performance, indicating the models are not independent.","kind":"statistical","quote":"The bagging experiment demonstrates that there is no significant improvement, regardless of whether the models are fine‑tuned or not."},{"finding":"The independence condition of Condorcet’s theorem is violated, suggesting similar decision‑making patterns across models of different sizes.","kind":"theoretical","quote":"Since the first condition is met, the failure to observe improvement suggests that the second condition is not fulfilled."}],"reported_performance":{"strategy":"sentiment signal equity strategy","sharpe":null,"sharpe_benchmark":null,"annual_return_pct":null,"total_return_pct":null,"max_drawdown_pct":null},"out_of_sample":{"method":"holdout","headline_out_of_sample":true}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":18,"citations":{"cited_by_in_corpus":1,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"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://huggingface.co/datasets/baptle/financial_headlines_market_based","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":"2026-11-14T18:03:35Z"},"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/"}