Bitcoin / Cryptocurrency Market - Trend Regression

Simple trend regressions on bitcoin price data + total cryptocurrency and altcoin market capitalizations.

Latest Readme Update: 21 Jan. 2018

Bitcoin Price Data

Exponential Regression

source("config/Config.R")

Sys.setenv(tz="UTC")
btc.price <- FetchBTCInfo(param           = "market-price",   
                          data.identifier = "btc.close", 
                          date.start      = "2011-01-01")

par(mfrow=c(1, 1))
SimpleLogTrendRegression(data = btc.price, 
                         data.identifier = "BTC-price",
                         regression.type = "exponential",
                         nr.future=120, plot.2sd.log = TRUE, plot.2sd.levels = FALSE)

Loess Regression

SimpleLogTrendRegression(data = btc.price, 
                         data.identifier = "BTC-price",
                         regression.type = "loess",
                         nr.future=120, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Logarithmic Regression (on log data)

SimpleLogTrendRegression(data            = btc.price,
                         data.identifier = "BTC-price",
                         regression.type = "logarithmic",
                         nr.future=120, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Cryptocurrency Market Capitalization (Total)

Note: The analysis below uses weekly historical data that was scraped from coinmarketcap.com and only goes back to 2013-04-28. Suggestions for better data sources are appreciated..

Exponential Regression

cryptocurrency.market.cap <- FetchCryptocurrencyMarketCapitalizations()
SimpleLogTrendRegression(data = cryptocurrency.market.cap$total,
                         data.identifier = "Total Cap", data.frequency = "weekly",
                         regression.type = "exponential",
                         nr.future       = 12, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Loess Regression

SimpleLogTrendRegression(data = cryptocurrency.market.cap$total,
                         data.identifier = "Total Cap", data.frequency = "weekly",
                         regression.type = "loess",
                         nr.future       = 12, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Logarithmic Regression (on log data)

SimpleLogTrendRegression(data = cryptocurrency.market.cap$total,
                         data.identifier = "Total Cap", data.frequency = "weekly",
                         regression.type = "logarithmic",
                         nr.future       = 12, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Cryptocurrency Market Capitalization (Altcoins)

Exponential Regression

SimpleLogTrendRegression(data = cryptocurrency.market.cap$alts,
                         data.identifier = "Alt Cap", data.frequency = "weekly",
                         regression.type = "exponential",
                         nr.future       = 12, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Loess Regression

SimpleLogTrendRegression(data = cryptocurrency.market.cap$alts,
                         data.identifier = "Alt Cap", data.frequency = "weekly",
                         regression.type = "loess",
                         nr.future       = 12, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Logarithmic Regression (on log data)

SimpleLogTrendRegression(data = cryptocurrency.market.cap$alts,
                         data.identifier = "Alt cap", data.frequency = "weekly",
                         regression.type = "logarithmic",
                         nr.future       = 12, plot.2sd.log = TRUE, plot.2sd.levels = TRUE)

Donations

If you find this software useful and/or you would like to see additional extensions, feel free to donate some crypto:

• BTC: 1QHtZLZ15Cmj4FPr5h5exDjYciBDhh7mzA
• LTC: LhKf6MQ7LY1k8YMaAq9z3APz8kVyFX3L2M
• ETH: 0x8E44D7C96896f2e0Cd5a6CC1A2e6a3716B85B479
• DASH: Xvicgp3ga3sczHtLqt3ekt7fQ62G9KaKNB

Licensing

Copyright 2017 Essential Data Science Consulting ltd. (EssentialQuant.com / [email protected]). This software is copyrighted under the MIT license: View added LICENSE file.