Joyce Liu

Papers

Research & publications

Can AI Do Financial Research? LLM-Guided Hypothesis Discovery in Asset Pricing

Huan Liu, Miao Liu, Zhizhe Liu, Danqing Mei · April 2026

We study whether an AI research agent can autonomously execute the hypothesis discovery loop in empirical asset pricing. Placing an LLM inside a human-designed research environment — a symbolic language of interpretable accounting formulas, an automated validation layer, and a fixed empirical evaluation pipeline — the agent iteratively proposes, tests, and revises signals across economic themes. Across eight theme pairs and seven generations, 280 candidates are evaluated; a small set survives multivariate horse races, modern asset pricing tests, and novelty checks against 209 published anomalies, carrying genuinely incremental information.

SSRN →

Human-AI Collaboration in Corporate Valuation: Experimental Evidence with a Valuation AI Agent

Huan Liu, Miao Liu, Zhizhe Liu, Danqing Mei · December 2025

We study how AI can facilitate human soft information production in corporate valuation. Using a retrieval-augmented AI agent embedded in an advanced business course, we find a non-monotonic effect: dashboards substantially increase soft information production, whereas adding AI-generated valuations induces anchoring and yields smaller gains. FSA knowledge amplifies dashboard benefits but does not eliminate crowding out.

SSRN →