FinanceHarness: Autonomous Financial Deep Research Framework
Summary
This paper introduces FinanceHarness, a framework for end-to-end automated financial deep research powered by LLM agents, along with FinanceGym, a verifiable point-in-time benchmark. Expert validation shows an 82% pass rate, while leading models score below 40%, and FinanceHarness improves open-weight backbone performance from 25.3% to 32.4%.
View Cached Full Text
Cached at: 07/31/26, 10:03 AM
# FinanceHarness: Autonomous Financial Deep Research Framework Source: [https://arxiv.org/abs/2607.27853](https://arxiv.org/abs/2607.27853) Authors:[Yijia Xiao](https://arxiv.org/search/cs?searchtype=author&query=Xiao,+Y),[Rujun Han](https://arxiv.org/search/cs?searchtype=author&query=Han,+R),[Yanfei Chen](https://arxiv.org/search/cs?searchtype=author&query=Chen,+Y),[Zifeng Wang](https://arxiv.org/search/cs?searchtype=author&query=Wang,+Z),[Ke Jiang](https://arxiv.org/search/cs?searchtype=author&query=Jiang,+K),[Zhongying CuiZhu](https://arxiv.org/search/cs?searchtype=author&query=CuiZhu,+Z),[Vishy Tirumalashetty](https://arxiv.org/search/cs?searchtype=author&query=Tirumalashetty,+V),[Wei Wang](https://arxiv.org/search/cs?searchtype=author&query=Wang,+W),[Burak Gokturk](https://arxiv.org/search/cs?searchtype=author&query=Gokturk,+B),[Tomas Pfister](https://arxiv.org/search/cs?searchtype=author&query=Pfister,+T),[Chen\-Yu Lee](https://arxiv.org/search/cs?searchtype=author&query=Lee,+C) [View PDF](https://arxiv.org/pdf/2607.27853) > Abstract:Powered by advances in LLMs and autonomous agents, deep research has become one of the most widely adopted agentic products\. However, most deep research systems write general\-purpose reports, which are inadequate for financial deep research\. Financial research demands specialized knowledge to analyze historical patterns and forecast upcoming events\. Automating financial deep research therefore requires both a layered harness to drive the research agent and a verifiable, point\-in\-time benchmark that prevents leakage of future information\. We present FinanceHarness, a harness that runs finance\-oriented tools and practitioner\-guided workflows, automating financial deep research end to end: environment and data construction, the agent execution loop, and reward modeling\. We further propose FinanceGym, comprising thesis\-driven research questions and rubrics that combine pre\-cutoff and post\-cutoff criteria\. Professional expert validation yields an 82% pass rate\. Even leading LLMs and agents score below 40% on the rubrics, showing that FinanceGym is challenging and leaves substantial headroom\. With the same open\-weight backbone, FinanceHarness improves the overall rubric score from 25\.3% to 32\.4%\. FinanceHarness is available at[this https URL](https://github.com/Yijia-Xiao/FinanceHarness)\. ## Submission history From: Yijia Xiao \[[view email](https://arxiv.org/show-email/c6d4bb62/2607.27853)\] **\[v1\]**Thu, 30 Jul 2026 08:32:04 UTC \(991 KB\)
Similar Articles
Harness Self-Evolution for Long-Horizon Terminal Agents
The paper presents RiseHarness, a framework from Tencent and multiple universities that iteratively improves agent harnesses for long-horizon terminal tasks while keeping model weights fixed, analyzing successful and failed runs to progressively integrate better execution mechanisms.
Dynamic Harness Search: Building Multi-Agent Systems Per-Query via Prediction
The paper introduces SHIFT, a framework that builds multi-agent harnesses per-query using a learned local LLM architect and value function combined with Monte Carlo tree search, achieving ~80% mean accuracy across 9,193 tasks and outperforming 17 baselines while reducing execution token costs by 32% in its cheaper mode.
HarnessX: A Composable, Adaptive, and Evolvable Agent Harness Foundry
HarnessX is a foundry for composable, adaptive, and evolvable AI agent harnesses that uses compositional primitives and trace-driven evolution to improve agent performance. Across five benchmarks, it achieves an average gain of +14.5% (up to +44.0%), demonstrating that runtime interface evolution is a complementary lever to model scaling.
HarnessForge: Joint Harness and Policy Evolution for Adaptive Agent Systems
HarnessForge proposes a meta-adaptive framework for evolving LLM agent systems by jointly optimizing the execution harness and reasoning policy, achieving consistent improvements on Qwen3 backbones across five benchmarks.
Self-Harness: Harnesses That Improve Themselves
Self-Harness introduces a new paradigm where LLM-based agents iteratively improve their own operating harness by mining model-specific weaknesses, proposing harness modifications, and validating them through regression testing, achieving substantial performance gains on Terminal-Bench-2.0 across multiple base models.