BankerToolBench: Evaluating AI Agents in End-to-End Investment Banking Workflows

2026-04-13

AI Summary

Existing AI benchmarks struggle to evaluate real-world performance in high-value, domain-specific workflows. This work introduces the first end-to-end investment banking evaluation benchmark grounded in authentic scenarios from 502 frontline practitioners, requiring AI agents to operate virtual data rooms, invoke industry tools—including market data platforms and SEC databases—and produce deliverables in multiple formats (Excel, PowerPoint, PDF/Word). The benchmark incorporates over 100 automated scoring criteria defined by senior bankers to assess client usability of outputs. Evaluation of nine state-of-the-art models reveals that even the best-performing model (GPT-5.4) fails to meet nearly half of these criteria and achieves 0% client readiness, exposing critical bottlenecks such as cross-artifact consistency.

Technology Category

• Humans and AI: AI for Accessibility
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models
• Machine Learning: Evaluation and Analysis

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Existing AI benchmarks lack the fidelity to assess economically meaningful progress on professional workflows. To evaluate frontier AI agents in a high-value, labor-intensive profession, we introduce BankerToolBench (BTB): an open-source benchmark of end-to-end analytical workflows routinely performed by junior investment bankers. To develop an ecologically valid benchmark grounded in representative work environments, we collaborated with 502 investment bankers from leading firms. BTB requires agents to execute senior banker requests by navigating data rooms, using industry tools (market data platform, SEC filings database), and generating multi-file deliverables--including Excel financial models, PowerPoint pitch decks, and PDF/Word reports. Completing a BTB task takes bankers up to 21 hours, underscoring the economic stakes of successfully delegating this work to AI. BTB enables automated evaluation of any LLM or agent, scoring deliverables against 100+ rubric criteria defined by veteran investment bankers to capture stakeholder utility. Testing 9 frontier models, we find that even the best-performing model (GPT-5.4) fails nearly half of the rubric criteria and bankers rate 0% of its outputs as client-ready. Our failure analysis reveals key obstacles (such as breakdowns in cross-artifact consistency) and improvement directions for agentic AI in high-stakes professional workflows.

Problem

• AI benchmarking
• investment banking
• professional workflows
• ecological validity
• agent evaluation

Innovation

• AI agent evaluation
• investment banking workflows
• ecologically valid benchmark
• multi-artifact consistency
• professional automation