BuySideFlow: A Point-in-Time Benchmark for LLM Agents in Investment Research

A new arXiv paper (2611.12966v1) introduces BuySideFlow, a point-in-time benchmark of 408 practitioner-derived Chinese-language investment-research tasks spanning equities, funds, fixed income and macroeconomics, drawn from a fixed snapshot of a licensed production financial database. Across nine agent configurations, result accuracy ranged from 16.2% to 36.8%, and an expert audit of 900 runs on 100 shared tasks found point-in-time violations in 193 of the 380 runs that passed result checks (50.8%). The authors conclude that output agreement alone does not establish temporal validity for LLM agents generating SQL and SQL-Python workflows in investment research.

arXiv:2611.12966v1 Announce Type: new Abstract: Large language models LLMs can generate SQL and SQL--Python workflows for investment research, but ensuring point-in-time PIT validity remains challenging. Agents must enforce disclosure cutoffs and select the appropriate historical states and record versions throughout data retrieval and financial analysis. We introduce \textsc{BuySideFlow}, a benchmark of 408 practitioner-derived Chinese-language tasks spanning equities, funds, fixed income, and macroeconomics over a fixed snapshot from a licensed production financial database. We provide executable references and a taxonomy of temporal errors, pairing result checks with expert audits of executed workflows. We evaluate nine agent configurations on all tasks, with result accuracy ranging from 16.2\% to 36.8\%. An audit of 900 runs on 100 shared tasks confirms PIT violations in 193 of the 380 runs that pass result checks 50.8\% , showing that output agreement alone does not establish temporal validity.