{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,30]],"date-time":"2026-10-30T07:30:32Z","timestamp":1793345432574,"version":"4.5.0"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":0,"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>Hybrid quantum-classical machine learning is increasingly tested in financial settings, yet evaluations rarely make it possible to attribute observed behavior to the quantum components themselves. We study one concrete decision problem: choosing a binary portfolio allocation under a budget constraint, where the risk term is estimated from generated joint credit-spread stress scenarios. We build a modular pipeline, SynQubi, with two quantum components in isolated, swappable roles: a Quantum Circuit Born Machine (QCBM) that generates tail scenarios, benchmarked against empirical, Gaussian, and Student-\ud835\udc61 copulas, and a QAOA optimizer, whose proposals are certified against exact classical solvers. Across seeds, on non-degenerate instances QAOA matches classical references on solution quality but is markedly worse on constraint satisfaction, returning budget-feasible allocations in roughly half of samples against the null model's all. Our main finding is methodological: certification overturned five of our own results that ordinary inspection had passed, and the errors pointed in different directions (two flattered the quantum components, three did not). Such errors cannot be fixed by discounting favorable results, only by a mechanical classical reference. A small pilot on LLM-generated trading-strategy code shows the same pattern in an LLM setting: a harness of mechanical validity checks found that every failure was a missing import, not a misunderstanding of the strategy. We establish no quantum advantage. CCS Concepts \u2022 Computing methodologies \u2192 Intelligent agents; \u2022 Applied computing \u2192 Economics.<\/jats:p>","DOI":"10.2139\/ssrn.7666979","type":"posted-content","created":{"date-parts":[[2026,10,30]],"date-time":"2026-10-30T06:31:03Z","timestamp":1793341863000},"source":"Crossref","is-referenced-by-count":0,"title":["Errors With No Common Sign: Certified Evaluation of Quantum and LLM Components in Financial Pipelines"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0009-0001-4190-4343","authenticated-orcid":true,"given":"Chevin","family":"Jeon","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Byungkwon","family":"Jeon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-4660-1316","authenticated-orcid":true,"given":"Michelle","family":"Jeon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","container-title":[],"original-title":[],"deposited":{"date-parts":[[2026,10,30]],"date-time":"2026-10-30T06:31:03Z","timestamp":1793341863000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=7666979"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026]]},"references-count":0,"URL":"https:\/\/doi.org\/10.2139\/ssrn.7666979","relation":{},"subject":[],"published":{"date-parts":[[2026]]},"subtype":"preprint"}}