RiskLabAI

RiskLabAI is a Julia library for quantitative finance, financial machine learning, and causal factor analysis. It provides research-oriented implementations of methods associated with Marcos López de Prado's Advances in Financial Machine Learning, Machine Learning for Asset Managers, and Causal Factor Investing.

The RiskLabAI 1.1.0 candidate preserves the 57-name causal-factor-analysis contract established in 1.0.0 and adds 30 paper-derived names, producing an 87-name causal API. The companion RiskLabAI.py package independently implements the same concepts. This parity statement applies to the causal API, not to every Julia module.

What is included

• Causal factor analysis - constrained minimum-variance allocation, factor-mirage and allocation-misspecification diagnostics, graphical identification and factor roles, deterministic structural-model evaluation, treatment-effect formulas, search-adjusted false discovery, specification experiments, and evidence records for the seven-stage causal-factor protocol
• Financial data structures - tick, volume, dollar, imbalance, run, and time bars
• Market features - entropy, microstructure, structural-break, and feature- importance utilities
• Portfolio and clustering methods - hierarchical risk parity, nested clustered optimization, hedging, correlation clustering, and silhouette diagnostics
• Backtest analytics - Sharpe-ratio inference, probability of backtest overfitting, strategy risk, multiple-testing corrections, bet sizing, and Ornstein-Uhlenbeck trading rules
• Validation - K-fold, purged, combinatorial-purged, and walk-forward validation, plus grid and random search
• PDE methods - a native Deep-BSDE solver using Lux, Optimisers, and Zygote

Compatibility

RiskLabAI 1.1.0 supports Julia 1.10.12 LTS and Julia 1.12.7. The complete tested policy and dependency details are in docs/compatibility.md.

Installation

To install the repository version directly from GitHub:

using Pkg
Pkg.add(url = "https://github.com/RiskLabAI/RiskLabAI.jl")

For a version available through Julia's General registry, use:

using Pkg
Pkg.add("RiskLabAI")

The Deep-BSDE solver and its numerical dependencies are installed with the package.

Causal-factor quick start

using RiskLabAI
using RiskLabAI.CausalFactorAnalysis
using LinearAlgebra: Diagonal

covariance = Matrix(Diagonal([1.0, 2.0, 4.0]))
factor_exposures = [1.0 0.0; 0.0 1.0; 1.0 1.0]
target_exposures = [0.0, 1.0]

weights = minimum_variance_factor_weights(
    covariance,
    factor_exposures,
    target_exposures,
)
@assert isapprox(weights, [-2.0 / 7.0, 5.0 / 7.0, 2.0 / 7.0])

effect = average_treatment_effect(3.5, 1.25)
@assert effect == 2.25

dag = CausalDAG(
    ("T", "U", "Y"),
    (("U", "T"), ("U", "Y"), ("T", "Y")),
    ("T", "U", "Y"),
)
@assert check_backdoor_adjustment_set(dag, "T", "Y", ("U",)).admissible

The complete deterministic example is examples/causal_factor_analysis_quickstart.jl. The causal API and its limits are documented in docs/causal_factor_analysis.md.

Development

git clone https://github.com/RiskLabAI/RiskLabAI.jl
cd RiskLabAI.jl
julia --project=. -e 'using Pkg; Pkg.instantiate(); Pkg.test()'

Please branch from main, keep changes focused, and include tests for behavior changes.

Scope

RiskLabAI is research software, not investment advice. Graph routines evaluate criteria on a caller-supplied directed acyclic graph; they do not discover or certify that graph. Protocol records validate declared evidence structures; they do not prove that empirical assumptions are true.

License

RiskLabAI is distributed under the BSD 3-Clause License.