Augmenting Fundamental Analysis with Large Language Models: A RAG-Based System for Generating Investor Briefs

2026-07-10 · Source: Computation and Language · Field: Finance & Economics — Capital Markets & Investment Management, FinTech & Digital Financial Services · Depth: Advanced, quick

Summary

A study investigated the application of Large Language Models (LLMs) to enhance fundamental analysis of companies, utilizing a Retrieval-Augmented Generation (RAG) system. This system preprocessed company reports, macroeconomic data such as GDP and inflation changes, and U.S. SEC filings from EDGAR to provide comprehensive context. The processed data was then sent via API to the gpt-4o model for brief generation. An exemplar investor knowledge document, based on Kitchin cycles, was also incorporated to enrich the analysis. The system analyzed data for 9 companies over a 4-week period, automatically generating investor briefs. These briefs were subsequently distributed to nine individual investors for evaluation, assessing the approach's usefulness in data analysis and potential for augmenting traditional methods.

Key takeaway

For financial analysts seeking to scale their fundamental analysis, integrating RAG-based LLM systems like the one using gpt-4o offers a path to automate investor brief generation. You can significantly reduce manual data aggregation from SEC filings and macroeconomic sources. Consider piloting such a system to produce initial drafts, allowing your team to focus on deeper qualitative insights and strategic recommendations, rather than routine report compilation.

Key insights

LLMs with RAG can automate investor brief generation from diverse financial and macroeconomic data.

Principles

• Combine company reports with macroeconomic data.
• Integrate expert knowledge like Kitchin cycles.
• Use RAG for context-aware LLM responses.

Method

Preprocess company reports, macroeconomic data, and SEC filings. Send to gpt-4o via API in a RAG regime. Generate automatic investor briefs.

In practice

• Automate brief creation for multiple companies.
• Evaluate LLM-generated content with target users.
• Incorporate Kitchin cycles for market context.

Topics

• Large Language Models
• Retrieval-Augmented Generation
• Fundamental Analysis
• Investor Briefs
• Financial Data Processing
• GPT-4o

Best for: NLP Engineer, Research Scientist, AI Scientist, Data Scientist, AI Engineer

Related on AIssential

• Augmenting Fundamental Analysis with Large Language Models: A RAG-Based System for Generating Investor Briefs — A RAG-based LLM system can automate and deepen fundamental investment analysis by synthesizing diverse data and generating tailored investor briefs.
• Retrieval-Augmented Large Language Models as Components of Cognitive Computing architecture for Regulatory Knowledge Management — RAG transforms local LLMs into reliable cognitive computing components for regulatory knowledge management.
• Fin-RATE: A Real-world Financial Analytics and Tracking Evaluation Benchmark for LLMs on SEC Filings — LLMs struggle with multi-document, multi-entity, and temporal financial reasoning, revealing critical gaps in current evaluation.
• Repository-Level Solidity Code Generation with Large Language Models: From Prompting to Fine-Tuning — High-quality domain data and supervised fine-tuning are crucial for reliable LLM-generated Solidity smart contracts.
• Vectorless RAG - Local Financial RAG Without Vector Database | Tree-Based Indexing with Ollama — Vectorless RAG uses document structure and LLMs for retrieval, eliminating vector databases for structured data.
• Time Series Augmented Generation for Financial Applications — Tool-augmented LLMs can achieve high accuracy and low hallucination in complex financial time-series analysis.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Computation and Language.