Agentic Portfolio Construction: A Multi-Agent Architecture for LLM-Driven Financial Asset Allocation
Bibliographic record
Abstract
Accurately curating portfolios and managing asset allocation is an essential and complex task in financial risk management. The emergence of large language models (LLMs) has helped resolve this task to a good extent. Recent literature have demonstrated that LLMs alone remain insufficient. We introduce a multi-agent portfolio construction architecture that integrates LLMs with an agentic orchestration architecture to enhance financial asset allocation workflows. With LangGraph, the system orchestrates interactions between LLM-based agents and functional nodes, enabling adaptive decision-making. The pipeline begins with parsing natural language investment objectives, followed by real-time data retrieval, financial metric computation, and portfolio generation tailored to the user’s risk profile and constraints. The results not only addresses risk-aware investing but also that the final output includes validation and a human-readable commentary. This approach addresses key deficiencies in standalone LLMs, such as lack of real-time awareness, by coordinating modular agents that specialize in tasks like data retrieval, analysis, and reporting. The architecture demonstrates how multi-agent LLM systems can perform complex, high-stakes financial tasks with adaptability, transparency, and robustness, marking a promising direction for AI-driven portfolio management.1
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".