▸ Working Papers & Original Research
Research
Archive
Original research bridging academic rigor and industry practice in Agentic AI, quantitative risk, and financial regulation.
▸ Agentic AI
Autonomous Risk Agents Under Market Stress: A Simulation Framework
We develop a multi-agent simulation framework to evaluate the behavior of LLM-based risk management agents during market stress episodes. Agents are benchmarked against historical drawdown periods including 2008, 2020, and 2022.
▸ Systemic Risk
Contagion Propagation in Agent-Based Financial Networks with Heterogeneous AI Actors
This paper models systemic contagion when market participants deploy heterogeneous agentic systems. We find that correlated agent architectures amplify tail risk by a factor of 2.3x relative to diverse decision-making populations.
▸ Regulation
SupTech 3.0: Large Language Models in Supervisory Surveillance
Regulators are piloting LLM-based surveillance tools for anomaly detection in trading data. We audit three central bank pilot programs and propose a governance framework for autonomous supervisory agents.
▸ NLP & Alt Data
Earnings Call Parsing at Scale: Agentic Information Extraction for Quantitative Signals
We evaluate six LLM architectures on a corpus of 40,000 earnings call transcripts, measuring signal-to-noise in extracting forward guidance, sentiment, and uncertainty markers for systematic trading.
▸ Risk Models
Interpretability Requirements for Autonomous Credit Decisioning Under ECOA
The Equal Credit Opportunity Act imposes explainability obligations that may conflict with frontier AI model architectures. This paper proposes a compliance taxonomy and audit protocol for agentic credit systems.
▸ Market Microstructure
Latency Arbitrage in the Age of Agentic Market Makers
We examine how autonomous LLM-based market makers interact with traditional HFT strategies at millisecond granularity. The results suggest new microstructure equilibria emerge when agent reaction functions are language-mediated.
▸ Call for Papers
Submit Research
RiskAICenter welcomes working papers, empirical studies, and theoretical contributions at the intersection of Agentic AI and financial risk. Selected papers will be presented at the Ski Summit and published in the proceedings.
Submission deadline: Septemmber 155, 2026
Submit Abstract →
▸ Agentic AI systems in portfolio management and risk
▸ LLM-based supervisory and compliance tools
▸ Agent-based models of systemic risk
▸ Explainability and governance of autonomous financial AI
▸ Alternative data pipelines and NLP for finance
▸ Reinforcement learning in trading and execution
▸ Central bank digital currencies and agentic payments