{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T15:45:06Z","timestamp":1792079106137,"version":"4.4.0"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":43,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,1,1]],"date-time":"2026-01-01T00:00:00Z","timestamp":1767225600000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/www.uspto.gov\/ip-policy\/copyright-policy\/copyright-basics"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>Does a client&amp;apos;s expressed distress change chatbot investment advice? In a preregistered experiment, Claude Opus 5.5, Gemini 3.8 Flash, and GPT-6 Sol answered 5,040 structured allocation requests. In 4,320, three hypothetical investors described a loss, an upcoming withdrawal, or both, in neutral or distressed words over identical financial facts. After a loss, every request restated an unchanged risk tolerance. Distress cut the share of allocations above a stock\u2013bond volatility reference matched to stated tolerance from 12.2% to 2.8% (-9.4 percentage points; 97.5% CI: -11.1 to -7.7), beyond the preregistered \u00b15-point equivalence band. The direction holds under alternative thresholds and covariances, although the decrease shrinks to 2.4\u20137.2 points. In an exploratory breakdown, the decrease came from the loss-only state (36.5% to 8.5%). All models accepted the task without asking for clarification. Coded retained-risk warnings ranged from 0.9% to 74.9% of explanations by the model. Whether warnings track risk within identical requests remains unresolved: two models barely varied their warnings, and the third model&amp;apos;s descriptive interval admits both tracking and none. Two models warned more often when their allocations were supplied back unchanged than when they chose them. Taken together, expressed feeling changes the portfolio itself, not just the explanation.<\/jats:p>","DOI":"10.2139\/ssrn.7615736","type":"posted-content","created":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T14:47:11Z","timestamp":1792075631000},"source":"Crossref","is-referenced-by-count":0,"title":["Expressed distress and portfolio risk in LLM investment advice"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7062-0032","authenticated-orcid":true,"given":"John","family":"Garcia","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","reference":[{"key":"ref1","doi-asserted-by":"crossref","first-page":"122","DOI":"10.1016\/j.jfineco.2020.05.009","article-title":"Testing the effectiveness of consumer financial disclosure: Experimental evidence from savings accounts","volume":"141","author":"P D Adams","year":"2021","journal-title":"J. Financ. Econ"},{"key":"ref2","doi-asserted-by":"crossref","DOI":"10.3390\/fintech5020054","article-title":"Generative AI as an investment advisor: Same client, different advice","volume":"5","author":"N Agliata","year":"2026","journal-title":"FinTech"},{"key":"ref3","author":"G Bahrami","year":"2026","journal-title":"Judge for yourself: Confidence beats accuracy in AI financial advice"},{"key":"ref4","article-title":"Inducing state anxiety in LLM agents reproduces human-like biases in consumer decision-making. npj","volume":"2","author":"Z Ben-Zion","year":"2026","journal-title":"Artif. Intell"},{"key":"ref5","doi-asserted-by":"crossref","DOI":"10.1016\/j.jbef.2020.100275","article-title":"Robo-advisors: A substitute for human financial advice?","volume":"25","author":"L Brenner","year":"2020","journal-title":"J. Behav. Exp. Financ"},{"key":"ref6","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2025.108923","article-title":"Multifaceted variability in LLM-driven stock recommendations","volume":"86","author":"S Chon","year":"2025","journal-title":"Finance Res. Lett"},{"key":"ref7","author":"T Choukhmane","year":"2026","journal-title":"AI financial advice: Supply, demand, and life cycle implications"},{"key":"ref8","doi-asserted-by":"crossref","first-page":"860","DOI":"10.1257\/aer.20131314","article-title":"Evidence for countercyclical risk aversion: An experiment with financial professionals","volume":"105","author":"A Cohn","year":"2015","journal-title":"Am. Econ. Rev"},{"key":"ref9","author":"J A Cookson","year":"2026","journal-title":"LLMs and systematic measurement error"},{"key":"ref10","doi-asserted-by":"crossref","first-page":"1983","DOI":"10.1093\/rfs\/hhz014","article-title":"The promises and pitfalls of robo-advising","volume":"32","author":"F D'acunto","year":"2019","journal-title":"Rev. Financ. Stud"},{"key":"ref11","doi-asserted-by":"crossref","first-page":"67","DOI":"10.1086\/466756","article-title":"Free competition and the optimal amount of fraud","volume":"16","author":"M R Darby","year":"1973","journal-title":"J. Law Econ"},{"key":"ref12","doi-asserted-by":"crossref","first-page":"233","DOI":"10.1086\/700735","article-title":"The market for financial adviser misconduct","volume":"127","author":"M Egan","year":"2019","journal-title":"J. Polit. Econ"},{"key":"ref13","author":"C Fieberg","year":"2025","journal-title":"Using large language models for financial advice"},{"key":"ref14","doi-asserted-by":"crossref","first-page":"1441","DOI":"10.1111\/jofi.12514","article-title":"Retail financial advice: Does one size fit all?","volume":"72","author":"S Foerster","year":"2017","journal-title":"J. Finance"},{"key":"ref15","doi-asserted-by":"crossref","first-page":"91","DOI":"10.1111\/jofi.12188","article-title":"Money doctors","volume":"70","author":"N Gennaioli","year":"2015","journal-title":"J. Finance"},{"key":"ref16","doi-asserted-by":"crossref","first-page":"403","DOI":"10.1016\/j.jfineco.2018.02.007","article-title":"Time varying risk aversion","volume":"128","author":"L Guiso","year":"2018","journal-title":"J. Financ. Econ"},{"key":"ref17","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2025.108102","article-title":"How good are LLMs in risk profiling?","volume":"85","author":"T Hens","year":"2025","journal-title":"Financ. Res. Lett"},{"key":"ref18","doi-asserted-by":"crossref","first-page":"60","DOI":"10.1016\/j.jbankfin.2012.08.007","article-title":"Individual investor perceptions and behavior during the financial crisis","volume":"37","author":"A O I Hoffmann","year":"2013","journal-title":"J. Bank. Financ"},{"key":"ref19","article-title":"Human feedback is not gold standard","author":"T Hosking","year":"2024","journal-title":"Proceedings of the Twelfth International Conference on Learning Representations (ICLR 2024)"},{"key":"ref20","doi-asserted-by":"crossref","first-page":"2086","DOI":"10.1257\/aer.20140386","article-title":"The realization effect: Risk-taking after realized versus paper losses","volume":"106","author":"A Imas","year":"2016","journal-title":"Am. Econ. Rev"},{"key":"ref21","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1093\/cesifo\/ifx001","article-title":"The economics of credence goods -a survey of recent lab and field experiments","volume":"63","author":"R Kerschbamer","year":"2017","journal-title":"CESifo Econ. Stud"},{"key":"ref22","doi-asserted-by":"crossref","first-page":"94","DOI":"10.1016\/j.jbef.2018.03.001","article-title":"Countercyclical risk aversion: Beyond financial professionals","volume":"18","author":"C K\u00f6nig-Kersting","year":"2018","journal-title":"J. Behav. Exp. Financ"},{"key":"ref23","doi-asserted-by":"crossref","first-page":"605","DOI":"10.1017\/S0022109011000123","article-title":"The influence of affect on beliefs, preferences, and financial decisions","volume":"46","author":"C M Kuhnen","year":"2011","journal-title":"J. Financ. Quant. Anal"},{"key":"ref24","doi-asserted-by":"crossref","first-page":"146","DOI":"10.1037\/0022-3514.81.1.146","article-title":"Fear, anger, and risk","volume":"81","author":"J S Lerner","year":"2001","journal-title":"J. Pers. Soc. Psychol"},{"key":"ref25","doi-asserted-by":"crossref","first-page":"799","DOI":"10.1146\/annurev-psych-010213-115043","article-title":"Emotion and decision making","volume":"66","author":"J S Lerner","year":"2015","journal-title":"Annu. Rev. Psychol"},{"key":"ref26","doi-asserted-by":"crossref","first-page":"587","DOI":"10.1111\/jofi.12995","article-title":"The misguided beliefs of financial advisors","volume":"76","author":"J T Linnainmaa","year":"2021","journal-title":"J. Finance"},{"key":"ref27","doi-asserted-by":"crossref","first-page":"391","DOI":"10.1146\/annurev-economics-080213-041341","article-title":"Disclosure: Psychology changes everything","volume":"6","author":"G Loewenstein","year":"2014","journal-title":"Annu. Rev. Econ"},{"key":"ref28","doi-asserted-by":"crossref","first-page":"267","DOI":"10.1037\/0033-2909.127.2.267","article-title":"Risk as feelings","volume":"127","author":"G F Loewenstein","year":"2001","journal-title":"Psychol. Bull"},{"key":"ref29","author":"X Martin","year":"2026","journal-title":"Sycophancy or information extraction? Evidence from realworld human-GAI interactions"},{"key":"ref30","article-title":"AI meets DIY: The impact of human intervention on AIassisted investing","volume":"6803678","author":"A Moss","year":"2026"},{"key":"ref31","author":"S Mullainathan","year":"2012","journal-title":"The market for financial advice: An audit study"},{"key":"ref32","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2023.104898","article-title":"Does ChatGPT provide better advice than robo-advisors?","volume":"60","author":"A Oehler","year":"2024","journal-title":"Financ. Res. Lett"},{"key":"ref33","doi-asserted-by":"crossref","first-page":"56","DOI":"10.1006\/obhd.1999.2838","article-title":"All negative moods are not equal: Motivational influences of anxiety and sadness on decision making","volume":"79","author":"R Raghunathan","year":"1999","journal-title":"Organ. Behav. Hum. Decis. Process"},{"key":"ref34","article-title":"One size fits none","author":"J Ross","year":"2026","journal-title":"Heuristic collapse in LLM investment advice"},{"key":"ref35","doi-asserted-by":"crossref","DOI":"10.1016\/j.jfineco.2024.103869","article-title":"The diversification and welfare effects of robo-advising","volume":"157","author":"A G Rossi","year":"2024","journal-title":"J. Financ. Econ"},{"key":"ref36","doi-asserted-by":"crossref","DOI":"10.1016\/j.frl.2025.107517","article-title":"Stock portfolio selection based on risk appetite: Evidence from ChatGPT","volume":"82","author":"C J Schneider","year":"2025","journal-title":"Financ. Res. Lett"},{"key":"ref37","doi-asserted-by":"crossref","first-page":"513","DOI":"10.1037\/0022-3514.45.3.513","article-title":"Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states","volume":"45","author":"N Schwarz","year":"1983","journal-title":"J. Pers. Soc. Psychol"},{"key":"ref38","article-title":"Towards understanding sycophancy in language models","author":"M Sharma","year":"2024","journal-title":"Proceedings of the Twelfth International Conference on Learning Representations"},{"key":"ref39","author":"C Shin","year":"2026","journal-title":"The effect of emotional context on large language models' endorsement of premature decisions: Comparing emotional vulnerability across six commercial models"},{"key":"ref40","author":"J Tait","year":"2026","journal-title":"Managing LLM bias in investing: From detection to mitigation"},{"key":"ref41","doi-asserted-by":"crossref","first-page":"643","DOI":"10.1287\/mnsc.36.6.643","article-title":"Gambling with the house money and trying to break even: The effects of prior outcomes on risky choice","volume":"36","author":"R H Thaler","year":"1990","journal-title":"Manage. Sci"},{"key":"ref42","doi-asserted-by":"crossref","DOI":"10.1371\/journal.pone.0325459","article-title":"Biased echoes: Large language models reinforce investment biases and increase the portfolio risks of private investors","volume":"20","author":"P Winder","year":"2025","journal-title":"PLOS ONE"},{"key":"ref43","author":"Z Zhao","year":"2026","journal-title":"The price of agreement: Measuring LLM sycophancy in agentic financial applications"}],"container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,15]],"date-time":"2026-10-15T14:50:33Z","timestamp":1792075833000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7615736"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":43,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7615736","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}