FinCARE: Financial Causal Analysis with Reasoning and Evidence

2025/10/23 by Alejandro Michel, Abhinav Arun, Michel, Alejandro +5 · 2 voices · 1 citation
Business, Management and Accounting · Computer Science · Decision Sciences · Economics, Econometrics and Finance · #Artificial Intelligence (cs.AI) #Computational Finance (q-fin.CP) #Explainable Artificial Intelligence (XAI) #FOS: Computer and information sciences #FOS: Economics and business #Financial Distress and Bankruptcy Prediction #Stock Market Forecasting Methods #cs.AI #q-fin.CP

paper · pdf · doi:10.48550/arxiv.2510.20221

openalex publication_date 2025/10/23 · arxiv published 2025/10/23 · arxiv updated 2025/10/23 · openalex created_date 2025/10/25 · openalex updated_date 2026/07/28

Abstract

Portfolio managers rely on correlation-based analysis and heuristic methods that fail to capture true causal relationships driving performance. We present a hybrid framework that integrates statistical causal discovery algorithms with domain knowledge from two complementary sources: a financial knowledge graph extracted from SEC 10-K filings and large language model reasoning. Our approach systematically enhances three representative causal discovery paradigms, constraint-based (PC), score-based (GES), and continuous optimization (NOTEARS), by encoding knowledge graph constraints algorithmically and leveraging LLM conceptual reasoning for hypothesis generation. Evaluated on a synthetic financial dataset of 500 firms across 18 variables, our KG+LLM-enhanced methods demonstrate consistent improvements across all three algorithms: PC (F1: 0.622 vs. 0.459 baseline, +36%), GES (F1: 0.735 vs. 0.367, +100%), and NOTEARS (F1: 0.759 vs. 0.163, +366%). The framework enables reliable scenario analysis with mean absolute error of 0.003610 for counterfactual predictions and perfect directional accuracy for intervention effects. It also addresses critical limitations of existing methods by grounding statistical discoveries in financial domain expertise while maintaining empirical validation, providing portfolio managers with the causal foundation necessary for proactive risk management and strategic decision-making in dynamic market environments.

Citations

• FinReflectKG: Agentic Construction and Evaluation of Financial Knowledge Graphs
• FinKario: Event-Enhanced Automated Construction of Financial Knowledge Graph
• Unveiling Causal Reasoning in Large Language Models: Reality or Mirage?
• CausalStock: Deep End-to-end Causal Discovery for News-driven Stock Movement Prediction
• Causal Agent based on Large Language Model
• On the Reliability of Large Language Models for Causal Discovery
• Fusing LLMs and KGs for Formal Causal Reasoning behind Financial Risk Contagion
• Multi-Agent Causal Discovery Using Large Language Models
• Hybrid Top-Down Global Causal Discovery with Local Search for Linear and Nonlinear Additive Noise Models
• ALCM: Autonomous LLM-Augmented Causal Discovery Framework
• Can Large Language Models Infer Causation from Correlation?
• Review of Causal Discovery Methods Based on Graphical Models
• DAGs with NO TEARS: Continuous Optimization for Structure Learning
• Pearl's Calculus of Intervention Is Complete
• A five-factor asset pricing model

Cited by

• Rethinking Retrieval: From Traditional Retrieval Augmented Generation to Agentic and Non-Vector Reasoning Systems in the Financial Domain for Large Language Models

Discussions

• Alejandro Michel, Abhinav Arun, Bhaskarjit Sarmah, Stefano Pasquali: FinCARE: Financial Causal Analysis with Reasoning and Evidence https://arxiv.org/abs/2510.20221 https://arxiv.org/pdf/2510.20221 ht [bsky, 0 points, 0 comments]
• FinCARE: Financial Causal Analysis with Reasoning and Evidence arxiv.org/abs/2510.20221 #QuantLinkADay [bsky, 0 points, 0 comments]

Related

• Causal Inference in Finance: An Expertise-Driven Model for Instrument Variables Identification and Interpretation
• Causality-Inspired Models for Financial Time Series Forecasting
• Fusing LLMs and KGs for Formal Causal Reasoning behind Financial Risk Contagion
• FinKario: Event-Enhanced Automated Construction of Financial Knowledge Graph
• Fino1: On the Transferability of Reasoning-Enhanced LLMs and Reinforcement Learning to Finance