Automated Market Intelligence: A Multi-Agent Approach with MCP

The sheer volume and velocity of information in financial markets create significant challenges for timely and accurate analysis. This paper presents a multi-agent system that uses relation extraction to derive actionable intelligence from financial news, corporate press releases, and market filings. The proposed Agentic AI system combines four agents: (i) a machine learning agent for sentiment analysis (logistic regression), (ii) a corporate profile agent for baseline fact-checking (which relies on named entity recognition), (iii) a narrative consistency agent (using LLM prompt engineering), and (iv) a realtime market data analyzer that extracts relational triplets for claim verification. The system is orchestrated via the Model Context Protocol (MCP), offering shared context and live learning across components. Results demonstrate that the multi-agent ensemble achieves 95.3% accuracy with an F1 score of 0.964 in identifying verifiable market claims, significantly outperforming individual agents and traditional approaches. The weighted aggregation method, mathematically derived from individual agent misclassification rates, proves superior to algorithmic threshold optimization. The modular architecture makes the system easily scalable while maintaining details of the analytical processes.