Skip to content

About

Example code of simple things one can do with our open-source asset pricing data

Resources

Stars

60 stars

Watchers

3 watching

Forks

Repository files navigation

CrossSectionDemos

Example code of simple things one can do with Open Source Asset Pricing data, created with our open source code.

Each file is an independent R script that downloads what it needs. OpenAP data come from the OpenAP Hugging Face dataset through the OpenSourceAP.DownloadR package. No account or token is needed, except for the two scripts that use CRSP data from WRDS.

Setup

install.packages(c("tidyverse", "gridExtra", "knitr", "OpenSourceAP.DownloadR"))

The scripts need OpenSourceAP.DownloadR 0.2.0 or later, which reads from Hugging Face. Check with packageVersion("OpenSourceAP.DownloadR").

Scripts use the latest data release by default. To use another, change OpenAP$new() to e.g. OpenAP$new('2024_10'). Releases from 2020_05 through 2025_10 are available.

The WRDS scripts read credentials from the environment variables WRDS_USER and WRDS_PASS or from ~/.pgpass, and prompt otherwise.

List of Scripts

Scripts are listed from lightest / fastest (top) to heaviest / slowest (bottom).

File Description
plot_anomaly.R Plots selected long-short returns along with sample end and publication dates
openap_vs_frenchweb.R Compares BM and BMdec long-short returns to Ken French's value portfolios
mclean_pontiff_main.R Replicates the fact that returns are mostly there out-of-sample (McLean and Pontiff 2016)
correlation_demo.R Correlations, principal components, and a dendrogram of long-short returns
make_pi_from_anomalies.R Estimates $\pi$ from anomaly returns, using the CLT and the mean absolute deviation of the standard normal
old_vs_new_returns_check.R Compares portfolio returns from two releases
daily_covid_demo.R Plots daily long-short returns (all signals) during the March 2020 crash
ff1993_style_implementation.R Builds a Fama-French 1993 style factor from a firm characteristic and CRSP (WRDS), compares to Ken French's HML and the release
dl_signals_add_crsp.R Creates the data for machine learning: all signed characteristics plus CRSP Price, Size, STreversal (WRDS, about 20 GB of RAM)

Details

  • Run each script from this folder, e.g. Rscript plot_anomaly.R.
  • Output and intermediate files go in temp/, which each script creates.

About

Example code of simple things one can do with our open-source asset pricing data

Resources

Stars

60 stars

Watchers

3 watching

Forks

Releases

Packages

Contributors

Languages