Crypto Currency Research Data from Coin-Market Cap without Survivorship Bias

Tue, 02 Aug 2022 00:00:00 +0000

In the past two years an ever-growing number of academic researchers has been investigating the market for cryptocurrencies (CC), often concentrating on the few largest ones (Brauneis and Mestel 2018; Bouri, Gupta, and Roubaud 2018; Corbet et al. 2018). However, in almost all studies only the surviving cryptocurrencies are considered and thereby all results are subject to survivorship bias.

Cryptocurrency data that is not subject to survivorship bias is hard to obtain, therefore I have created my own package crypto2 that is now available from CRAN.

Historical Cryptocurrency Prices for Active and Delisted Tokens!

This is a modification of the original crypto package by jesse vent. It is entirely set up to use means from the tidyverse and provides tibbles with all data available via the web-api of coinmarketcap.com. It does not require an API key but in turn only provides information that is also available through the website of coinmarketcap.com.

It allows the user to retrieve

crypto_listings() a list of all coins that were historically listed on CMC (main dataset to avoid delisting bias) according to the CMC API documentation
crypto_list() a list of all coins that are listed as either being active, delisted or untracked according to the CMC API documentation
crypto_info() a list of all information available for all available coins according to the CMC API documentation
crypto_history() the most powerful function of this package that allows to download the entire available history for all coins covered by CMC according to the CMC API documentation
fiat_list() a mapping of all fiat currencies (plus precious metals) available via the CMC WEB API
exchange_list() a list of all exchanges available as either being active, delisted or untracked according to the CMC API documentation
exchange_info() a list of all information available for all given exchanges according to the CMC API documentation

Update

Since version 1.4.4 a new function crypto_listings() was introduced that retrieves new/latest/historical listings and listing information at CMC. Additionally some aspects of the other functions have been reworked. We noticed that finalWait = TRUE does not seem to be necessary at the moment, as well as sleep can be set to ‘0’ seconds. If you experience strange behavior this might be due to the the api sending back strange (old) results. In this case let sleep = 60 (the default) and finalWait = TRUE (the default).

Since version 1.4.0 the package has been reworked to retrieve as many assets as possible with one api call, as there is a new “feature” introduced by CMC to send back the initially requested data for each api call within 60 seconds. So one needs to wait 60s before calling the api again. Additionally, since version v1.4.3 the package allows for a data interval larger than daily (e.g. ‘2d’ or ‘7d’/‘weekly’)

Installation

You can install crypto2 from CRAN with

install.packages("crypto2")

or directly from github with:

# install.packages("devtools")
devtools::install_github("sstoeckl/crypto2")

Package Contribution

The package provides API free and efficient access to all information from https://coinmarketcap.com that is also available through their website. It uses a variety of modification and web-scraping tools from the tidyverse (especially purrr).

As this provides access not only to active coins but also to those that have now been delisted and also those that are categorized as untracked, including historical pricing information, this package provides a valid basis for any Asset Pricing Studies based on crypto currencies that require survivorship-bias-free information. In addition to that, the package maintainer is currently working on also providing delisting returns (similarly to CRSP for stocks) to also eliminate the delisting bias.

Package Usage

First we load the crypto2-package and download the set of active coins from https://coinmarketcap.com (additionally one could load delisted coins with only_Active=FALSE as well as untracked coins with add_untracked=TRUE).

library(crypto2)
library(dplyr)

##
## Attache Paket: 'dplyr'
## Die folgenden Objekte sind maskiert von 'package:stats':
##
## filter, lag
## Die folgenden Objekte sind maskiert von 'package:base':
##
## intersect, setdiff, setequal, union

# List all active coins
coins <- crypto_list(only_active=TRUE)

Next we download information on the first three coins from that list.

# retrieve information for all (the first 3) of those coins
coin_info <- crypto_info(coins, limit=3, finalWait=FALSE)

## ❯ Scraping crypto info
##
## Scraping https://web-api.coinmarketcap.com/v1/cryptocurrency/info?id=1,2,3 with 65 characters!
## ❯ Processing crypto info
##
## ❯ Sleep for 60s before finishing to not have next function call end up with this data!
##

# and give the first two lines of information per coin
coin_info

## # A tibble: 3 × 19
## id name symbol categ…¹ descr…² slug logo subre…³ notice date_…⁴ twitt…⁵
## * <int> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr> <chr>
## 1 1 Bitco… BTC coin "## Wh… bitc… http… bitcoin "" 2013-0… ""
## 2 2 Litec… LTC coin "## Wh… lite… http… liteco… "" 2013-0… "Litec…
## 3 3 Namec… NMC coin "Namec… name… http… nameco… "" 2013-0… "Namec…
## # … with 8 more variables: is_hidden <int>, date_launched <lgl>,
## # self_reported_circulating_supply <lgl>, self_reported_market_cap <lgl>,
## # tags <list>, self_reported_tags <lgl>, urls <list>, platform <list>, and
## # abbreviated variable names ¹category, ²description, ³subreddit,
## # ⁴date_added, ⁵twitter_username
## # ℹ Use `colnames()` to see all variable names

In a next step we show the logos of the three coins as provided by https://coinmarketcap.com.

In addition we show tags provided by https://coinmarketcap.com.

coin_info %>% select(slug,tags) %>% tidyr::unnest(tags) %>% group_by(slug) %>% slice(1,n())

## # A tibble: 6 × 2
## # Groups: slug [3]
## slug tags
## <chr> <chr>
## 1 bitcoin mineable
## 2 bitcoin paradigm-portfolio
## 3 litecoin mineable
## 4 litecoin bnb-chain
## 5 namecoin mineable
## 6 namecoin platform

Additionally: Here are some urls pertaining to these coins as provided by https://coinmarketcap.com.

coin_info %>% select(slug,urls) %>% tidyr::unnest(urls) %>% filter(name %in% c("reddit","twitter"))

## # A tibble: 5 × 3
## slug name url
## <chr> <chr> <chr>
## 1 bitcoin reddit https://reddit.com/r/bitcoin
## 2 litecoin twitter https://twitter.com/LitecoinProject
## 3 litecoin reddit https://reddit.com/r/litecoin
## 4 namecoin twitter https://twitter.com/Namecoin
## 5 namecoin reddit https://reddit.com/r/namecoin

In a next step we download time series data for these coins.

# retrieve historical data for all (the first 3) of them
coin_hist <- crypto_history(coins, limit=3, start_date="20210101", end_date="20210105", finalWait=FALSE)

## ❯ Scraping historical crypto data
##
## ❯ Processing historical crypto data
##

# and give the first two times of information per coin
coin_hist %>% group_by(slug) %>% slice(1:2)

## # A tibble: 6 × 16
## # Groups: slug [3]
## timestamp id slug name symbol ref_cur open high low
## <dttm> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <dbl>
## 1 2021-01-01 23:59:59 1 bitcoin Bitc… BTC USD 2.90e+4 2.96e+4 2.88e+4
## 2 2021-01-02 23:59:59 1 bitcoin Bitc… BTC USD 2.94e+4 3.32e+4 2.91e+4
## 3 2021-01-01 23:59:59 2 liteco… Lite… LTC USD 1.25e+2 1.33e+2 1.23e+2
## 4 2021-01-02 23:59:59 2 liteco… Lite… LTC USD 1.26e+2 1.40e+2 1.24e+2
## 5 2021-01-01 23:59:59 3 nameco… Name… NMC USD 4.39e-1 4.63e-1 4.32e-1
## 6 2021-01-02 23:59:59 3 nameco… Name… NMC USD 4.51e-1 5.10e-1 4.15e-1
## # … with 7 more variables: close <dbl>, volume <dbl>, market_cap <dbl>,
## # time_open <dttm>, time_close <dttm>, time_high <dttm>, time_low <dttm>
## # ℹ Use `colnames()` to see all variable names

Similarly, we could download the same data on a monthly basis.

# retrieve historical data for all (the first 3) of them
coin_hist_m <- crypto_history(coins, limit=3, start_date="20210101", end_date="20210501", interval ="monthly", finalWait=FALSE)

## ❯ Scraping historical crypto data
##
## ❯ Processing historical crypto data
##

# and give the first two times of information per coin
coin_hist_m %>% group_by(slug) %>% slice(1:2)

## # A tibble: 6 × 16
## # Groups: slug [3]
## timestamp id slug name symbol ref_cur open high low
## <dttm> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <dbl>
## 1 2021-01-01 23:59:59 1 bitcoin Bitc… BTC USD 2.90e+4 2.96e+4 2.88e+4
## 2 2021-02-01 23:59:59 1 bitcoin Bitc… BTC USD 3.31e+4 3.46e+4 3.24e+4
## 3 2021-01-01 23:59:59 2 liteco… Lite… LTC USD 1.25e+2 1.33e+2 1.23e+2
## 4 2021-02-01 23:59:59 2 liteco… Lite… LTC USD 1.30e+2 1.36e+2 1.26e+2
## 5 2021-01-01 23:59:59 3 nameco… Name… NMC USD 4.39e-1 4.63e-1 4.32e-1
## 6 2021-02-01 23:59:59 3 nameco… Name… NMC USD 7.82e-1 8.05e-1 7.48e-1
## # … with 7 more variables: close <dbl>, volume <dbl>, market_cap <dbl>,
## # time_open <dttm>, time_close <dttm>, time_high <dttm>, time_low <dttm>
## # ℹ Use `colnames()` to see all variable names

Alternatively, we could determine the price of these coins in other currencies. A list of such currencies is available as fiat_list()

fiats <- fiat_list()
fiats

## # A tibble: 93 × 4
## id name sign symbol
## <int> <chr> <chr> <chr>
## 1 2781 United States Dollar $ USD
## 2 2782 Australian Dollar $ AUD
## 3 2783 Brazilian Real R$ BRL
## 4 2784 Canadian Dollar $ CAD
## 5 2785 Swiss Franc Fr CHF
## 6 2786 Chilean Peso $ CLP
## 7 2787 Chinese Yuan ¥ CNY
## 8 2788 Czech Koruna Kč CZK
## 9 2789 Danish Krone kr DKK
## 10 2790 Euro € EUR
## # … with 83 more rows
## # ℹ Use `print(n = ...)` to see more rows

So we download the time series again depicting prices in terms of Bitcoin and Euro (note that multiple currencies can be given to convert, separated by “,”).

# retrieve historical data for all (the first 3) of them
coin_hist2 <- crypto_history(coins, convert="BTC,EUR", limit=3, start_date="20210101", end_date="20210105", finalWait=FALSE)

## ❯ Scraping historical crypto data
##
## ❯ Processing historical crypto data
##

# and give the first two times of information per coin
coin_hist2 %>% group_by(slug,ref_cur) %>% slice(1:2)

## # A tibble: 12 × 16
## # Groups: slug, ref_cur [6]
## timestamp id slug name symbol ref_cur open high low
## <dttm> <int> <chr> <chr> <chr> <chr> <dbl> <dbl> <dbl>
## 1 2021-01-01 23:59:43 1 bitco… Bitc… BTC BTC 1 e+0 1.00e+0 9.98e-1
## 2 2021-01-02 23:59:43 1 bitco… Bitc… BTC BTC 1 e+0 1.00e+0 9.99e-1
## 3 2021-01-01 23:59:06 1 bitco… Bitc… BTC EUR 2.37e+4 2.43e+4 2.36e+4
## 4 2021-01-02 23:59:06 1 bitco… Bitc… BTC EUR 2.42e+4 2.73e+4 2.40e+4
## 5 2021-01-01 23:59:43 2 litec… Lite… LTC BTC 4.30e-3 4.56e-3 4.27e-3
## 6 2021-01-02 23:59:43 2 litec… Lite… LTC BTC 4.30e-3 4.24e-3 4.23e-3
## 7 2021-01-01 23:59:06 2 litec… Lite… LTC EUR 1.02e+2 1.09e+2 1.01e+2
## 8 2021-01-02 23:59:06 2 litec… Lite… LTC EUR 1.04e+2 1.16e+2 1.02e+2
## 9 2021-01-01 23:59:43 3 namec… Name… NMC BTC 1.51e-5 1.58e-5 1.50e-5
## 10 2021-01-02 23:59:43 3 namec… Name… NMC BTC 1.54e-5 1.57e-5 1.31e-5
## 11 2021-01-01 23:59:06 3 namec… Name… NMC EUR 3.60e-1 3.80e-1 3.54e-1
## 12 2021-01-02 23:59:06 3 namec… Name… NMC EUR 3.71e-1 4.21e-1 3.41e-1
## # … with 7 more variables: close <dbl>, volume <dbl>, market_cap <dbl>,
## # time_open <dttm>, time_close <dttm>, time_high <dttm>, time_low <dttm>
## # ℹ Use `colnames()` to see all variable names

As a new features in version 1.4.4. we introduced the possibility to download historical listings and listing information (add quote = TRUE).

latest_listings <- crypto_listings(which="latest", limit=10, quote=TRUE, finalWait=FALSE)
latest_listings

## # A tibble: 10 × 23
## id name symbol slug self_…¹ self_…² tvl_r…³ last_upd…⁴ USD_p…⁵ USD_v…⁶
## <int> <chr> <chr> <chr> <int> <dbl> <lgl> <date> <dbl> <dbl>
## 1 1 Bitcoin BTC bitc… NA NA NA 2022-08-03 2.33e+4 2.56e10
## 2 52 XRP XRP xrp NA NA NA 2022-08-03 3.75e-1 1.09e 9
## 3 825 Tether USDT teth… NA NA NA 2022-08-03 1.00e+0 4.67e10
## 4 1027 Ethere… ETH ethe… NA NA NA 2022-08-03 1.64e+3 1.65e10
## 5 1839 BNB BNB bnb NA NA NA 2022-08-03 3.02e+2 2.13e 9
## 6 2010 Cardano ADA card… NA NA NA 2022-08-03 5.10e-1 5.47e 8
## 7 3408 USD Co… USDC usd-… NA NA NA 2022-08-03 1.00e+0 6.88e 9
## 8 4687 Binanc… BUSD bina… NA NA NA 2022-08-03 1.00e+0 5.29e 9
## 9 5426 Solana SOL sola… NA NA NA 2022-08-03 3.95e+1 2.31e 9
## 10 6636 Polkad… DOT polk… 9.05e8 7.48e9 NA 2022-08-03 8.27e+0 4.97e 8
## # … with 13 more variables: USD_volume_change_24h <dbl>,
## # USD_percent_change_1h <dbl>, USD_percent_change_24h <dbl>,
## # USD_percent_change_7d <dbl>, USD_percent_change_30d <dbl>,
## # USD_percent_change_60d <dbl>, USD_percent_change_90d <dbl>,
## # USD_market_cap <dbl>, USD_market_cap_dominance <dbl>,
## # USD_fully_diluted_market_cap <dbl>, USD_tvl <lgl>,
## # USD_market_cap_by_total_supply <dbl>, USD_last_updated <chr>, and …
## # ℹ Use `colnames()` to see all variable names

Last and least, one can get information on exchanges. For this download a list of active/inactive/untracked exchanges using exchange_list():

exchanges <- exchange_list(only_active=TRUE)
exchanges

## # A tibble: 499 × 6
## id name slug is_active first_historical_data last_histor…¹
## <int> <chr> <chr> <int> <date> <date>
## 1 16 Poloniex poloniex 1 2018-04-26 2022-08-03
## 2 21 BTCC btcc 1 2018-04-26 2022-08-03
## 3 22 Bittrex bittrex 1 2018-04-26 2022-08-03
## 4 24 Kraken kraken 1 2018-04-26 2022-08-03
## 5 34 Bittylicious bittylicious 1 2018-04-26 2022-08-03
## 6 36 CEX.IO cex-io 1 2018-04-26 2022-08-03
## 7 37 Bitfinex bitfinex 1 2018-04-26 2022-08-03
## 8 42 HitBTC hitbtc 1 2018-04-26 2022-08-03
## 9 50 EXMO exmo 1 2018-04-26 2022-08-03
## 10 61 Okcoin okcoin 1 2018-04-26 2022-08-03
## # … with 489 more rows, and abbreviated variable name ¹last_historical_data
## # ℹ Use `print(n = ...)` to see more rows

and then download information on “binance” and “kraken”:

ex_info <- exchange_info(exchanges %>% filter(slug %in% c('binance','kraken')), finalWait=FALSE)

## ❯ Scraping exchange info
##
## Scraping exchanges from https://web-api.coinmarketcap.com/v1/exchange/info?id=24,270 with 60 characters!
## ❯ Processing exchange info
##

ex_info

## # A tibble: 2 × 19
## id name slug descr…¹ notice logo type date_…² is_hi…³ is_re…⁴ maker…⁵
## * <int> <chr> <chr> <lgl> <chr> <chr> <chr> <chr> <int> <lgl> <dbl>
## 1 24 Kraken krak… NA "" http… "" 2011-0… 0 NA 0.02
## 2 270 Binance bina… NA "Bina… http… "" 2017-0… 0 NA 0.02
## # … with 8 more variables: taker_fee <dbl>, spot_volume_usd <dbl>,
## # spot_volume_last_updated <dttm>, weekly_visits <int>, tags <lgl>,
## # urls <list>, countries <lgl>, fiats <list>, and abbreviated variable names
## # ¹description, ²date_launched, ³is_hidden, ⁴is_redistributable, ⁵maker_fee
## # ℹ Use `colnames()` to see all variable names

Then we can access information on the fee structure,

ex_info %>% select(contains("fee"))

## # A tibble: 2 × 2
## maker_fee taker_fee
## <dbl> <dbl>
## 1 0.02 0.05
## 2 0.02 0.04

the amount of cryptocurrencies being traded (in USD)

ex_info %>% select(contains("spot"))

## # A tibble: 2 × 2
## spot_volume_usd spot_volume_last_updated
## <dbl> <dttm>
## 1 515861153. 2022-08-03 22:15:16
## 2 12884611026. 2022-08-03 22:15:16

or the fiat currencies allowed:

ex_info %>% select(slug,fiats) %>% tidyr::unnest(fiats)

## # A tibble: 53 × 2
## slug value
## <chr> <chr>
## 1 kraken USD
## 2 kraken EUR
## 3 kraken GBP
## 4 kraken CAD
## 5 kraken JPY
## 6 kraken CHF
## 7 kraken AUD
## 8 binance AED
## 9 binance ARS
## 10 binance AUD
## # … with 43 more rows
## # ℹ Use `print(n = ...)` to see more rows

Author/License

Sebastian Stöckl - Package Creator, Modifier & Maintainer - sstoeckl on github

This project is licensed under the MIT License - see the <license.md> file for details</license.md>

Acknowledgments

Thanks to the team at https://coinmarketcap.com for the great work they do, especially to Alice Liu (Research Lead) and Aaron K. for their support with regard to information on delistings.
Thanks to Jesse Vent for providing the (not fully research compatible) crypto-package that inspired this package.

References

Bouri, Elie, Rangan Gupta, and David Roubaud. 2018. “Herding Behaviour in Cryptocurrencies.” Finance Research Letters, July. https://doi.org/10.1016/j.frl.2018.07.008.

Brauneis, Alexander, and Roland Mestel. 2018. “Price Discovery of Cryptocurrencies: Bitcoin and Beyond.” Economics Letters 165: 58–61. https://doi.org/10.1016/j.econlet.2018.02.001.

Corbet, Shaen, Andrew Meegan, Charles Larkin, Brian Lucey, and Larisa Yarovaya. 2018. “Exploring the Dynamic Relationships Between Cryptocurrencies and Other Financial Assets.” Economics Letters 165 (April): 28–34. https://doi.org/10.1016/j.econlet.2018.01.004.

crypto2 | Sebastian Stöckl