FinCARE: Financial Causal Analysis with Reasoning and Evidence
This work addresses the need for more accurate causal analysis in finance for portfolio managers, offering a domain-specific enhancement to existing methods.
The paper tackles the problem of capturing true causal relationships in financial performance analysis by introducing a hybrid framework that integrates statistical causal discovery algorithms with domain knowledge from a financial knowledge graph and large language model reasoning, resulting in consistent improvements across three algorithms (e.g., F1 scores up to 0.759 vs. 0.163 baseline) and reliable scenario analysis with low error rates.
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.