Download Data
All files are updated weekly and can be downloaded from the Hugging Face dataset. The files are free for academic and non-commercial use. Please cite the project and Liu et al. (2022).
@article{liu2022common,
author = {Liu, Yukun and Tsyvinski, Aleh and Wu, Xi},
title = {Common Risk Factors in Cryptocurrency},
journal = {The Journal of Finance},
volume = {77},
number = {2},
pages = {1133--1177},
year = {2022},
doi = {10.1111/jofi.13119}
}
@misc{stoeckl2026opencrypto,
author = {Stoeckl, Sebastian and Pukrop, Moritz},
title = {Open Crypto Pricing},
year = {2026},
institution = {University of Liechtenstein},
url = {https://huggingface.co/datasets/sstoeckl/opencryptoassetpricing}
}Factor series
For the analysis we calculated two different factor series for each of the three factors (CMKT, CSIZE, CMOM) using two independent data sources (CoinMarketCap and CoinGecko). The investable series compute breakpoints on the investable universe (the recommended series — see Results); the all-coin series follow the original full-universe convention for comparison with Liu et al. (2022).
| Source | Universe | CSV | Parquet |
|---|---|---|---|
| CoinGecko | Investable (recommended) | factors_cg_investable.csv | .parquet |
| CoinGecko | All-coin (Liu-style) | factors_cg_all.csv | .parquet |
| CoinMarketCap | Investable | factors_cmc_investable.csv | .parquet |
| CoinMarketCap | All-coin (Liu-style) | factors_cmc_all.csv | .parquet |
The factors are computed with the following canonical specification: stablecoins excluded, sharp calendar, value-weighted, quintile size sort, two-week momentum, gap-corrected returns, investable-momentum signal.
Multiverse & analysis tables
The full specification multiverse and the cross-sectional evidence behind the Results:
| File | Contents |
|---|---|
| multiverse_results.csv | One row per specification (4,608 worlds): factor means + Fama-MacBeth pricing stats |
| factor_survival.csv | Long-short significance of the 10 characteristics, full vs investable |
| calendar_study.csv | Pricing by week-start calendar (investable world) |
| calendar_study_allcoin.csv | Pricing by week-start calendar (all-coin world) |
All data
To download all files at once, use the Hugging Face dataset repository: https://huggingface.co/datasets/sstoeckl/opencryptoassetpricing. With the following selection you can select the files you want to download directly from the site:
The URL is built directly from the Hugging Face dataset files.
Use in R / Python
Using R or Python, the files can be directly read from our website or from the Hugging Face dataset. The following example shows how to read the CoinGecko investable factors:
# direct from the site
f <- read.csv("https://opencryptopricing.com/data/factors_cg_investable.csv")
# or parquet via arrow
library(arrow)
f <- read_parquet("https://opencryptopricing.com/data/factors_cg_investable.parquet")import pandas as pd
f = pd.read_csv("https://opencryptopricing.com/data/factors_cg_investable.csv")The underlying coin data is retrieved with the crypto2 R package (CoinMarketCap and CoinGecko), so every series here is fully reproducible from source.
File format
| Column | Description |
|---|---|
week_start |
Week start date (YYYY-MM-DD), sharp-calendar convention |
CMKT |
Market excess return (value-weighted, minus 1-month T-bill) |
CSIZE |
Size factor (small minus big) |
CMOM |
Momentum factor (winners minus losers, two-week) |