GitHub

BayesianFactorZoo.jl

BayesianFactorZoo.jl is a Julia port of the R package 'BayesianFactorZoo' (CRAN link) implementing the econometric methods from the paper:

Bryzgalova, S., Huang, J., & Julliard, C. (2023). Bayesian solutions for the factor zoo: We just ran two quadrillion models. Journal of Finance, 78(1), 487–557. DOI: 10.1111/jofi.13197

For a more detailed function documentations please see the documentation of the R package Link to PDF

Overview

BayesianFactorZoo.jl provides a comprehensive framework for analyzing linear asset pricing models that is:

  • Simple and robust
  • Applicable to high-dimensional problems
  • Capable of handling both tradable and non-tradable factors

For a stand-alone model, the package delivers reliable price of risk estimates for tradable and non-tradable factors.

Installation

The package is registered in the General registry and so can be installed at the REPL with ] add BayesianFactorZoo or by running:

    using Pkg 
    Pkg.add("BayesianFactorZoo")

Alternatively you can install the latest dev version directly from this repository.

using Pkg
Pkg.add(url="http://github.com/eohne/BayesianFactorZoo.jl")

Quick Start

using BayesianFactorZoo

# Example with simulated data
t, k, N = 600, 3, 25  # time periods, factors, assets
f = randn(t, k)       # factor returns
R = randn(t, N)       # asset returns

# Perform Bayesian Fama-MacBeth regression
results_fm = BayesianFM(f, R, 10_000)

# Estimate SDF with normal prior
results_sdf = BayesianSDF(f, R; prior="Normal")

# Model selection with continuous spike-and-slab
results_ss = continuous_ss_sdf(f, R, 10_000)

Features

  • Bayesian Fama-MacBeth regression (BayesianFM)
  • Bayesian SDF estimation (BayesianSDF)
  • Model selection via spike-and-slab priors (continuous_ss_sdf, continuous_ss_sdf_v2)
  • Hypothesis testing (dirac_ss_sdf_pvalue)
  • GMM estimation (SDF_gmm)
  • Classical two-pass regression (TwoPassRegression)

Citation

If you use this package, please cite:

@article{bryzgalova2023bayesian,
  title={Bayesian solutions for the factor zoo: We just ran two quadrillion models},
  author={Bryzgalova, Svetlana and Huang, Jiantao and Julliard, Christian},
  journal={The Journal of Finance},
  volume={78},
  number={1},
  pages={487--557},
  year={2023}
}