{"paper_id":"2507.20957","title":"Your AI, Not Your View: The Bias of LLMs in Investment Analysis","authors":["Hoyoung Lee","Junhyuk Seo","Suhwan Park","Junhyeong Lee","Wonbin Ahn","Chanyeol Choi","Alejandro Lopez-Lira","Yongjae Lee"],"paper_date":"2025-07-28","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2507.20957v4","page_url":"https://thequant.space/flowcharts/your-ai-not-your-view-the-bias-of-llms-in-investment-analy/","scores":{"empirical_rigor":6,"math_complexity":4,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes a three-stage experimental framework to systematically elicit and quantify latent biases in LLMs for investment analysis.","quote":"To answer these questions, this study introduces a three-stage experimental framework designed to systematically elicit and verify LLM biases in investment analysis.","summary":"In finance, Large Language Models (LLMs) face frequent knowledge conflicts arising from discrepancies between their pre-trained parametric knowledge and real-time market data. These conflicts are especially problematic in real-world investment services, where a model’s inherent biases can misalign…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Evidence is generated by a single LLM (Gemini-2.5-Pro) and simplified to a numerical value.","type":"data","stated":"explicit","quote":"First, our evidence was generated by a single LLM and simplified to a numerical value, an approach that may not fully capture the biases of the generator model or the complexity of real-world information.","support":"partial"},{"assumption":"All evidence is generated (no external data) to avoid generation bias.","type":"model","stated":"explicit","quote":"To neutralize this potential generation bias and ensure that observed bias are not artifacts of context sourcing, our methodology exclusively uses generated evidence for all experimental conditions.","support":"full"},{"assumption":"Models are run at temperature τ=0.6 to balance determinism and creativity.","type":"model","stated":"explicit","quote":"All experiments were performed with the models configured at a temperature of 𝜏= 0.6, striking a balance between deterministic and creative response generation, and each experimental run was conducted three times to ensure the robustness of our results.","support":"full"},{"assumption":"Buy and sell evidence are presented in equal proportion with identical intensity (I=5%).","type":"data","stated":"explicit","quote":"Specifically, for every stock 𝑠, buy evidence (E (𝑠) buy) and sell evi-dence (E (𝑠) sell) are generated in an equal proportion, yielding a com-prehensive dataset of |E| = 3, 416 evidence. To further isolate bias, all evidence is engineered with a uniform linguistic structure and explicitly states a fixed intensity parameter 𝐼= 5%.","support":"full"},{"assumption":"Evidence order is randomized across trials to mitigate positional bias.","type":"execution","stated":"explicit","quote":"To quantify this elicited bias, the decision task is repeated 𝑁= 10 times for each stock, with the evidence order randomized in each trial to mitigate positional bias.","support":"full"}]},"datasets":{"status":"available","items":[{"name":"Curated set of 427 prominent stocks","provider":null,"kind":"fundamentals","access":"public","period":null,"universe":"S&P 500 constituents continuously listed over past five years","quote":"Accordingly, our investigation is confined to a curated set of 427 prominent stocks, denoted S = {𝑠1,𝑠2, . . . ,𝑠427}. These stocks were selected for their continuous listing in the S&P 500 index over the past five years.","support":"full"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"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":"unclear","findings":[{"finding":"Across most models, there is a tendency to prefer technology stocks, large‑cap stocks, and contrarian strategies.","kind":"statistical","quote":"Across most models, a tendency to prefer technology stocks, large-cap stocks, and contrarian strategies is observed.","support":"full"},{"finding":"Size bias scores decline as company size decreases, with statistically significant gaps between largest and smallest market‑cap quantiles.","kind":"statistical","quote":"Bias scores consistently decline as company size decreases. To statistically validate these observed trends, we performed independent-samples t-tests comparing the bias scores between the highest- and lowest-bias quantiles for each model (Table 2).","support":"full"},{"finding":"LLMs exhibit confirmation bias, clinging to initial judgments despite increasing counter‑evidence.","kind":"negative","quote":"These foundational biases often escalate into confirmation bias, causing models to cling to initial judgments even when faced with increasing counter-evidence. We demonstrate how these biases lead to a stub-born refusal to revise judgments when the model is challenged with stronger counter-evidence in terms of both volume and intensity.","support":"full"}],"reported_performance":null,"out_of_sample":{"method":"none","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[{"paper":{"paper_id":"2411.00420","title":"Evaluating Company-specific Biases in Financial Sentiment Analysis using Large Language Models","paper_date":"2024-11-01","authors":["Kei Nakagawa","Masanori Hirano","Yugo Fujimoto"],"empirical_rigor":7.5,"page_url":"https://thequant.space/flowcharts/evaluating-company-specific-biases-in-financial-sentiment-an/"},"question":"Do large language models show a company‑size sentiment bias in financial analysis?","this_paper_finding":"Size bias scores decline as company size decreases, with statistically significant gaps between largest and smallest market‑cap quantiles.","other_paper_finding":"Company-specific sentiment bias is present across multiple LLMs, with GPT-3.5 showing a negative evaluation of smaller firms when the company name is included.","other_paper_quote":"This suggests that GPT-3.5 tends to negatively evaluate smaller companies when the company name is included in the prompt.","note":"Both find that bias grows with size: GPT‑3.5 rates smaller firms more negatively, and bias scores drop as market‑cap decreases, with significant gaps between the largest and smallest quantiles.","confidence":0.804}],"contradicts":[],"qualifies":[],"compared_with":12,"citations":{"cited_by_in_corpus":10,"by_independent_authors":5,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":1,"discusses_findings":0,"mentions":3,"unclassified":6},"independent_papers":[{"paper":{"paper_id":"2510.05702","title":"Uncovering Representation Bias for Investment Decisions in Open-Source Large Language Models","paper_date":"2025-10-07","authors":["Fabrizio Dimino","Krati Saxena","Bhaskarjit Sarmah","Stefano Pasquali"],"empirical_rigor":7,"page_url":"https://thequant.space/flowcharts/uncovering-representation-bias-for-investment-decisions-in-o/"},"how":"extends","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":2,"extracted_at":"2026-11-14T19:48:39Z"},"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/"}