Algorithmic CAPM: Simulated Asset Returns for AI-Driven Portfolio Delegation

874.8 KB of code and data accompanies a paper developing an Algorithmic Capital Asset Pricing Model (CAPM). The simulation framework includes 500 assets, 200 investors, and 20 AI algorithms to study equilibrium pricing under AI-driven portfolio delegation. Authored by hanzun li and last updated on 2026-05-28, the dataset provides a benchmark for testing how algorithmic signals reshape factor pricing.

Use Cases

• Testing asset pricing models based on the described algorithmic beta factor.
• Simulating market equilibrium under conditions of AI-driven portfolio delegation.
• Benchmarking the performance of traditional CAPM versus algorithmic CAPM specifications.
• Analyzing the relationship between algorithmic consensus signals and equilibrium risk premia.

Strengths

• Includes simulation results for a defined scenario with 500 assets, 200 investors, and 20 AI algorithms.
• Provides quantitative model performance metrics, such as a cross-sectional R² of 0.0595 for the two-factor specification.
• Released under a permissive CC-BY-4.0 license for open use and modification.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The data is simulated for a theoretical model, not observed from real markets.

Provenance

Source

figshare

Collection Method

Simulation data generated for academic research on algorithmic asset pricing.

Freshness

Last updated 2026-05-28 09:29:53; freshness should be verified.